EXTENDED JOURNAL ABSTRACT

Transcription

1

2

3 EXTENDED JOURNAL ABSTRACT Targeted Journal: Accident Analysis and Prevention An Evaluation of Bicycle Awareness Zones Nicole Smart ABSTRACT A project was undertaken to review Bicycle Awareness Zones (BAZ). The project involved video analysis of the interaction between cyclists and motorists; a questionnaire survey of motorists and cyclists about their perceptions of BAZ; and crash data analysis of Brisbane cycle crashes. The video analysis showed that BAZ were effective on roads with up to AADT with minimal impact to the traffic flow in constrained width conditions, where BAZ are interspersed with bicycle lanes. The observed cyclists operating widths and offsets from parked cars have provided argument for changes to the BAZ symbol, its position and/ or size. The questionnaire survey showed motorists had a lack of understanding of the purpose of the BAZ symbols, whilst cyclists had a better understanding. Both motorists and cyclists agreed that BAZ make a safer road environment for cyclists than a road without any bicycle symbols. Based on the outcomes of the crash analysis, there is no evidence to suggest BAZ are statistically more dangerous than a bicycle lane. Introduction Bicycle Awareness Zones (BAZ) are an on- road bicycle treatment first introduced by Brisbane City Council as a trial in 1996/7 and are now also installed on Sunshine Coast Regional Council and Gold Coast City Council roads (Connell Wagner 2001). BAZ are based on a similar treatment, known as sharrows, which were developed by the City of Denver and are now being used across America (Yeates 2000). BAZ are an advisory bicycle treatment, rather than a dedicated bicycle facility. They are implemented on roads which, according to the Austroads Guide to Road Design or Manual of Uniform Traffic Control Devices, are too narrow for a designated bicycle lane. BAZ are defined by a yellow on- road bicycle symbol. They advise motorists of the possibility of onroad cyclists and provide guidance to cyclists on the position they should take on the road (Transport and Main Roads 2009).

4 In 2009, BAZ were included in the Queensland Department of Transport and Main Roads (TMR) Traffic and Road Use Management Manual (TRUM) which contains guidelines for the use of BAZ. A final year university project, under the direction of the TMR, reviewed the safety and effectiveness of Bicycle Awareness Zones. This project coincided with a TMR review of the current operational guidelines for the use of BAZ. BAZ were installed in the TRUM under the provision that the guidelines be reviewed in 24 months (Traffic Management Practices Committee 2011). The following sections describe the project elements used to evaluate BAZ and the results of each element. A summary of the safety and effectiveness of BAZ is presented and a list of the recommendations formulated from the results of each element. Methodologies The evaluation of BAZ was based on the findings of crash data analysis, video analysis and a questionnaire survey of motorists and cyclists. Video Analysis Video analysis was conducted at three Brisbane BAZ sites: Annerley Road, Riding Road and Kedron Brook Road. These sites are sufficiently used by cyclists and they have varying traffic volumes, parking lane width and parking turnover. This made the sites suitable for a comparison of cyclists and motorists behaviour between the sites. An overhead mounted video camera was erected at each site. From the videos of each site, the positions of the cyclists relative to passing vehicles, parked cars and the parking edge line were measured. In addition, any conflicts or crashes were noted. BAZ Questionnaire Survey The motoring public were approached at parks adjacent to roads with BAZ symbols and asked to participate in a survey that tested their understanding of on- road bicycle facility linemarking and their perception of its safety. A survey specifically for cyclists was ed to organizers of bicycle user groups in the Brisbane and Sunshine Coast Regions for distribution to their members. The cyclists survey included additional questions about past history of conflicts and crashes. Crash Data Analysis The crash data for all cyclist crashes occurring in Brisbane between 2004 and 2009 were obtained using WebCrash, an

5 online database containing crash information for crashes which occurred on Queensland s roads. There were 1100 bicycle crashes occurring in Brisbane in the 5 year period. However the majority of crashes occurred on roads without bicycle facilities and as such, only 100 crashes were analysed. Crash reports and site characteristics were assessed for each mid- block bicycle lane and BAZ crash to determine if there were any links between certain site characteristics and crashes. Any trends in BAZ crashes were compared to trends in bicycle lane crashes to determine if BAZ facilities were any more dangerous than dedicated bicycle lanes. Results Video Analysis The video analysis provided an indication of the operational space required by a cyclist on roads with varying parking widths. The required clearance to parked cars and passing vehicles could be used as a guide in the design of bicycle facilities. In constrained conditions, the operating envelope ranged from 1.58m in peak hour to 1.9m outside peak period. BAZ Symbol Cyclists were found to generally ride 3m from the kerb. If the centre of the symbol which is aligned with the parking edge line is supposed to represent the intended line of travel of the cyclist, it is suggested the centre of the BAZ symbol be positioned 3m from the kerb, except when the parking demand is low on narrow roads. The mean operating space of the cyclist even in constrained conditions is larger than the width of the current 1.1m wide symbol. If the intent of the symbol is to indicate the positioning of the cyclist, a new symbol may be warranted which should be closer to the minimum operating space in constrained conditions which was 1580mm. Cyclists Impact on Traffic Traffic was not disrupted on Riding Road and Kedron Brook Road which had peak hour volumes in the recorded direction of 400veh/h and 955veh/h respectively. On Riding Road, which had a 3m parking lane and a 3.3m traffic lane, the majority of motorists passed rather than followed the cyclists and as a result, queuing was minimal even in peak hour. On Kedron Brook Road, where the parking lane was 3.5m and the traffic lane was 3.3m, there were no disruptions to traffic since all motorists had sufficient width to pass the cyclist.

6 On Annerley Road which was the most constrained environment with a 2.5m parking lane and a 3.3m traffic lane and 1200veh/h ( AADT), the majority of motorists chose to follow rather than pass a cyclist. Queuing was observed in peak conditions, though the impact to the traffic stream was minimal due to the interspersion of BAZ and bicycle lanes. BAZ Questionnaire Survey The responses from the motorists indicated a lack of understanding of the purpose of BAZ. Of the three purposes of BAZ listed on the questionnaire, the purpose most known was that BAZ warn of the presence of cyclists, which was recognized by 55% of motorists. Understanding of bicycle lanes was far more acceptable with 90% recognizing that the white bicycle symbol designates a bicycle lane. The responses from cyclists indicated a greater understanding of bicycle facilities. As would be expected, cyclists commented that they feel safer riding in a bicycle lane than a BAZ. However, BAZ were still considered to be safe by cyclists. Crash Data Analysis Crash rate calculations, based on crashes per km per year of each type of facility, suggest that cycling in a BAZ is no more hazardous than cycling in a designated bicycle lane. Mid- block crashes are more likely to occur on BAZ due to greater probability of dooring but of the 12 midblock crashes associated with BAZ, only five dooring incidents were recorded. The majority of BAZ (70% of total) and bicycle lane (89% of total) crashes occurred at intersections which are known to be regions of high potential conflict (Forester 1994, p 97). Aside from the greater risk of a dooring incident (which was only minimal over five years), the available data suggests that BAZ don t have any obvious crash history that would lead to a conclusion that BAZ are unsafe to use. Conclusions video analysis demonstrated that BAZ can be effective on roads with up to AADT with minimal impact to the traffic flow. In constrained width conditions, the impact on the traffic stream can be minimized by interspersing BAZ with bicycle lanes; changes to the BAZ symbol, position and/ or size could be justified based on calculated operating envelopes and offsets from parked cars; questionnaire survey showed a general lack of understanding of the meaning of the BAZ symbols particularly by motorists but both

7 motorists and cyclists groups agreed that BAZ makes a safer road environment for cyclists than a road without any bicycle symbols at all; based on the outcomes of the crash analysis, there is no evidence to suggest BAZ are statistically more dangerous than a bicycle lane; bicycle lanes should always be used in preference to a BAZ where the carriageway width permits but BAZ have proven to be a useful and effective link in local government cycle networks. envelopes and perhaps the inclusion of tracking lines; other aspects of BAZ linemarking, such as the preferred position of the symbol relative to the edge line and the spacing of symbols, also need to be considered; educate motorists and cyclists about BAZ. References Connell Wagner 2001, Bicycle Brisbane Plan Review: Review of Bicycle- Friendly Zones, CT/ Revision 1, Brisbane. Recommendations BAZ are suitable for roads with traffic volumes up AADT but on narrow roads with high traffic volumes, consideration should be given to providing reasonable passing opportunities; installation of BAZ on a road needs consideration of parameters such as the available width, cyclist operating widths, the length of BAZ, passing opportunities, traffic volume and parking demand; consideration needs to be given for the adoption of the TMR (2009) standard 1530mm BAZ symbol to better reflect the operating Forester, J 1997, Bicycle Transportation: A Handbook for Cycling Transportation Engineers, The MIT Press, Massachusetts. Traffic Management Practices Committee 2011, Minutes from 4/2/2011, Transport and Main Roads, Brisbane. Transport and Main Roads 2009, Traffic and Road Use Management Manual: Bicycle Awareness Zones, Transport and Main Roads, Brisbane. Yeates, M 2000, Making space for cyclists by sharing the road Brisbane City Council s Bicycle Friendly Zone, paper presented to the National Road Safety Conference, Brisbane, 26 November 2000.

8 Griffith School of Engineering Griffith University 4001ENG Industrial Affiliates Program An Evaluation of Bicycle Awareness Zones Nicole Eileen Smart (s ) 10/6/2011 (Semester 1) Department of Transport and Main Roads Mark McDonald Dr. Kali Prasad Nepal A report submitted in partial fulfillment of the degree of Bachelor of Civil Engineering The copyright on this report is held by the author and/or the IAP Industry Partner. Permission has been granted to Griffith University to keep a reference copy of this report.

9 EXECUTIVE SUMMARY Bicycle Awareness Zones (BAZ) are an advisory bicycle treatment, rather than a dedicated bicycle lane, represented by yellow on- road bicycle symbols. The Queensland Department of Transport and Main Roads (TMR) Traffic and Road Use Management Manual (TRUM) provides guidelines for the use of BAZ. An IAP project was undertaken in conjunction with TMR to evaluate the safety and effectiveness of BAZ with the results to be considered in a scheduled review of the TRUM. Three elements of analysis were used to evaluate BAZ: video analysis of the interactions between motorists and cyclists; questionnaire survey of motorists and cyclists; and crash data analysis. Video analysis was conducted at three Brisbane BAZ sites. Video cameras recorded the interactions between motorists and cyclists and their positions on the road. BAZ were found to be effective on roads with up to AADT with minimal impact to the traffic flow in constrained width conditions where BAZ are interspersed with bicycle lanes. The operating envelopes and offsets from parked cars observed in the analysis have provided argument for changes to the BAZ symbol, position and/ or size. Separate questionnaires were developed for motorists and members of bicycle user groups. The results of the surveys showed motorists had a lack of understanding of the purpose of BAZ, whilst cyclists had a better understanding. Both motorists and cyclists agreed that BAZ makes a safer road environment for cyclists than a road without any bicycle symbols. Crashes involving cyclists on either a BAZ or bicycle lane were analysed to determine if BAZ are any more dangerous than dedicated bicycle lanes. Based on the outcomes of the crash analysis, there is no evidence to suggest BAZ are statistically more dangerous than a bicycle lane. A number of recommendations have been made based on the outcomes of the analysis. The main recommendations are as follows: BAZ are appropriate for use on roads with an AADT of 3000 or greater subject to consideration of parking conditions, operating envelopes and passing opportunities; consider changes to the BAZ symbol, placement and spacing; educate motorists and cyclists about BAZ. i

10 ACKNOWLEDGEMENTS Throughout the duration of this project, a number of contributions were made by others to ensure successful completion. The Department of Transport and Main Roads, in particular Mark McDonald and Robyn Davies, must be acknowledged for accepting me to take on the project and for providing the necessary resources. Firstly I would like to thank my industry supervisor, Mark McDonald. In addition to his continuous support and suggestions, Mark provided a significant amount of his time helping me use the Geographic Information System software, MapInfo. Mark also thoroughly reviewed all documents I prepared and provided helpful feedback. Robyn Davies also offered invaluable support and reviewed a number of documents, always providing feedback and ongoing suggestions. The expertise of Heidi Bushell, also from the Department of Transport and Main Roads, was a strong asset to the project. Heidi used her experience in developing surveys to ensure that the questions in the surveys produced for this project would prompt the required information from the respondents. Heidi formatted the surveys online and provided assistance in collating the results. It is quite possible that the video analysis wouldn t have been as successful without the enormous contribution made by Skyhigh Data. Skyhigh Data supplied and erected video cameras and supplied the video recordings at no cost. They also checked on the cameras and the road markings on a regular basis, even over the Easter break. Finally, the guidance and feedback provided by Dr. Kali Nepal, the Griffith University academic advisor for the project, must be recognized. ii

11 TABLE OF CONTENTS EXECUTIVE SUMMARY... I ACKNOWLEDGEMENTS... II 1 INTRODUCTION Background Project Justification Objectives Summary of Methodologies Structure of the Report LITERATURE REVIEW Relevant Standards and Guidelines Advisory Treatments Used Elsewhere Video Analysis Sharrow s Influence on the Position of Motorists and Cyclists BAZ EVALUATION APPROACHES Video Analysis BAZ Questionnaire Survey Crash Data Analysis VIDEO ANALYSIS Methodologies Characteristics and Locations of Video Sites Summary of Video Statistics Annerley Road Peak Period Annerley Road- Non Peak Kedron Brook Road Peak Period Kedron Brook Road- Non Peak Riding Road Peak Period Riding Road- Combined Period Validity of TRUM Guidelines iii

12 4.7.1 Cyclist Envelope BAZ Symbol (Alignment, Size and Symbol Profile) Annual Average Daily Traffic Amendments to Parking BAZ through Intersections Multilane Roads BAZ QUESTIONNAIRE SURVEY Methodology Survey of Motorists Motorists Survey Sample Size Motorists Awareness Motorists Understanding Motorists Behaviour Motorists Response to a Scenario Motorists Comments with Respect to BAZ Survey of Cyclists Cyclists Survey Sample Size Cyclist s Awareness Cyclists Understanding Cyclists Response Regarding the Safety of Bicycle Facilities Cyclists Response to Cycling Hazards Cyclists Response Regarding Conflicts Cyclists Response Regarding Crashes Cyclists Response Regarding Motorists Behaviour in BAZ Cyclists Response Regarding Motorists Behaviour in Bicycle Lanes Cyclists Response Regarding Issues with BAZ Recommendations Made by Cyclists Comparison between Motorists and Cyclists Responses Comparison with Connell Wagner Study Conclusions of Questionnaire Survey CRASH DATA ANALYSIS Methodology Crash Rate Calculations iv

13 6.3 Analysis of Mid- Block Crashes General Analysis of Mid- Block Crashes Detailed Assessment of Mid-Block Crashes BAZ Dooring Incidents AADT at Mid-Block Crash Locations Multiple Crashes at Mid-Block Locations Analysis of Intersection Crashes General Analysis of Intersection Crashes Multiple Crashes at Intersections Roundabouts Conclusions of Crash Data Analysis Data Collection by Road Authorities CONCLUSIONS General Video Analysis BAZ Questionnaire Survey Crash Data Analysis Summary v

14 8 RECOMMENDATIONS REFERENCES APPENDIX A: VIDEO ANALYSIS CODING APPENDIX B: SUGGESTED SYMBOL DESIGNS B MM X 1800MM SYMBOL, 3M OFFSET FROM KERB, WITH YELLOW TRACKING LINES B MM X 1800MM SYMBOL, 3M OFFSET FROM KERB, WITH ARROWS AS TRACKING LINES B MM X 2500MM SYMBOL, 3M OFFSET FROM KERB, WITH YELLOW TRACKING LINES B MM X 2500MM SYMBOL, 3M OFFSET FROM KERB, WITH ARROWS AS TRACKING LINES B MM X 2500MM SYMBOL, 3M OFFSET FROM KERB, NO TRACKING LINES APPENDIX C: TRUM GUIDELINES C.1 LANE CONFIGURATIONS WITHOUT PARKING C.2 ROAD CONFIGURATIONS WITH PARALLEL PARKING APPENDIX D: BAZ QUESTIONNAIRE SURVEYS D.1 MOTORISTS SURVEY D.2 CYCLISTS SURVEY APPENDIX E: SURVEY RESULTS E.1 MOTORISTS SURVEY RESULTS E.2 CYCLISTS SURVEY RESULTS APPENDIX F: MOTORISTS SURVEY COMMENTS F.1 POSITIVE ASPECTS OF BAZ F.2 ISSUES OR PROBLEMS WITH BAZ APPENDIX G: HAZARDS FACED BY CYCLISTS APPENDIX H: DESCRIPTIONS OF CYCLISTS CONFLICTS APPENDIX I: DESCRIPTIONS OF CYCLISTS CRASHES APPENDIX J: DCA CODING SYSTEM APPENDIX K: INTERSECTIONS WITH MULTIPLE CRASHES K.1 CRASHES AT INTERSECTIONS WITH A BAZ ON APPROACH K.2 CRASHES AT INTERSECTIONS WITH A BICYCLE LANE ON APPROACH 135 vi

15 K.3 CRASHES AT INTERSECTIONS WITH NO BICYCLE FACILITY ON APPROACH vii

16

17 1 INTRODUCTION 1.1 Background Cycle networks are comprised of off- road paths, on- road bicycle lanes, and wide kerbside lanes (Austroads 2011). In the 1990s, as Brisbane City Council tried to expand its on road facilities, it found that road widths and parking constraints made it increasingly difficult to install bike lanes whilst still maintaining parking. Council recognised the need to implement a solution which would provide a safe and convenient cycle network whilst still maintaining parking. This lead to trial projects of alternative facilities (Yeates 2000). After a failed trial of wide shared parking lanes, Brisbane City Council adopted the City of Denver s approach of shared use lanes, identifiable by pavement markings (Yeates 2000). Brisbane City Council s version of Denver s shared use lanes are called Bicycle Friendly Zones and they were introduced as a trial in 1996/7 (Connell Wagner 2001). Bicycle Friendly Zones were later renamed Bicycle Awareness Zones, heron to be referred to as BAZ. BAZ are now used on Gold Coast City Council and Sunshine Coast Regional Council roads. BAZ are an advisory bicycle treatment, rather than a dedicated bicycle facility. They are implemented on roads which, according to the Austroads Guide to Road Design or Manual of Uniform Traffic Control Devices, are too narrow for a designated bicycle lane. BAZ are defined by a yellow on- road bicycle symbol. They advise motorists of the possibility of onroad cyclists and provide guidance to cyclists on the position they should take on the road (Transport and Main Roads 2009). In 2002, representatives of a Traffic Management Practices Committee who provide advice on the operational practice and road safety measures on Queensland s road network, recommended that BAZ be included in the Traffic and Road Use Management Manual (TRUM) with a provision to be reviewed in 24 months (Traffic Management Practices Committee 2011). BAZ were included in the TRUM in The TRUM stipulates thresholds for the use of BAZ based on Austroads guidelines for formal bicycle facilities given in Austroads Guide to Traffic Engineering Part 14: Bicycles (Austroads 1999). At above 3000 vehicles per day, the guidelines recommend use of a formal bicycle lane rather than BAZ (Austroads 2011). However, a number of local governments would like to be able to use BAZ Nicole Eileen Smart 1

18 on roads with traffic volumes greater than this because of the width and parking constraints on many local roads. This IAP project conducted under the direction of the Department of Transport and Main Roads (TMR) will review the safety and effectiveness of Bicycle Awareness Zones. This project coincides with the Department s review of the current operational guidelines for the use of BAZ as stipulated in the TRUM. Results and findings from this project will be considered in the scheduled revision of the TRUM guidelines on BAZ which commenced in May 2011 (Traffic Management Practices Committee 2011). 1.2 Project Justification Apart from a small public survey on the effectiveness of BAZ which was undertaken by Connell Wagner in 2001 when BAZ were quite new, very little research has been undertaken on the safety or otherwise of BAZ. No research has been undertaken on the traffic thresholds stated in the TRUM, and whether installing BAZ on roads with an annual average daily traffic (AADT) greater than 3000 vehicles per day, has any implications. Using crash data, together with on- site observations, this project should establish relationships that could be used in TMR s review of the TRUM operational guidelines for BAZ. Currently the TRUM gives some guidelines on the appropriate usage of BAZ but there is some reluctance by engineers to install BAZ notwithstanding TRUM guidelines because, Main Roads deems that BAZ provides neither a safe nor attractive facility (Transport and Main Roads 2009). By clarifying operating parameters of cyclists on different road configurations and establishing links between crash history, bicycle facility and their locality characteristics, this project will provide relationships that designers can use to quantify the risks of using BAZ in different road environments. The information provided in this project should help establish if BAZ have any specific safety issues or whether they have a high incidence of conflicts and crashes. One of the recommendations in the TRUM was that local education campaigns be implemented to improve the education of motorists and cyclists with respect to BAZ (Transport and Main Roads 2009). The Department has no information on the implementation of such programs and hence the need for a current test of the public s knowledge. Random questionnaire surveys undertaken as part of this project will identify if 2 An Evaluation of Bicycle Awareness Zones

19 education campaigns are still required and will show if there has been a change in the public s understanding of BAZ since the 2001 Connell Wagner study. 1.3 Objectives To date, there has been minimal evaluation of the safety and effectiveness of the treatment. The research and findings of this project will seek to improve the understanding and operational implementation of BAZ, and could potentially lead to recommendations of a review of the traffic volume and other thresholds stipulated by the TRUM which provide guidance to the appropriateness of a BAZ at a particular site. 1.4 Summary of Methodologies Three project elements were used to evaluate BAZ: video analysis; a questionnaire survey of motorists and cyclists; and crash data analysis. Video analysis was conducted at three Brisbane BAZ sites to observe the interactions between motorists and cyclists and their positions on the road. A questionnaire survey was asked of motorists and cyclists to gain an understanding of their knowledge of and experiences with BAZ. Cyclist crashes occurring on roads with BAZ or bicycle lanes were analysed to determine if BAZ are any more dangerous than dedicated bicycle lanes. 1.5 Structure of the Report This report will outline the investigations and results of the project review of the safety and effectiveness of BAZ. Specifically, the following will be discussed: literature review of BAZ and sharrows; purpose of each project element being video analysis; questionnaire survey; and crash data analysis; methodologies of each project element; results and discussion of the results for each project element; recommendations. Nicole Eileen Smart 3

20 2 LITERATURE REVIEW 2.1 Relevant Standards and Guidelines The Traffic and Road Use Management Manual provides guidelines on the installation of BAZ and refers to the Manual of Uniform Traffic Control Devices for linemarking (Transport and Main Roads 2009). 2.2 Advisory Treatments Used Elsewhere New South Wales advisory bicycle treatment is called bicycle shoulder lanes. As with BAZ, bicycle shoulder lanes are installed on roads too narrow for a bicycle lane. They are typically used when there is high demand for parking. Bicycle shoulder lanes have a similar appearance to bicycle lanes, however they are narrower and the adjacent parking lane has a reduced width in comparison to bicycle lanes. The treatment also incorporates parking tee pavement markings to encourage vehicles to park as close to the kerb as possible. This feature is generally not included with BAZ treatments (RTA 2005). In the United Kingdom, an advisory treatment is used at locations where motorists are likely to encroach on a cycle lane if one existed. A coloured surface is used for the lane to alert motorists of cyclists in situations where motorists may block the lane, such as when motorists are queuing to join the main road. The advisory treatment generally has lane markings, though these may be dropped, leaving just the coloured surface and providing symbols and direction arrows, at junctions (Cycling England 2008). Some American cities use a similar treatment to BAZ called sharrows. Like BAZ, sharrows convey to motorists and cyclists that they must share the road on which they are travelling and clarify where the cyclists are expected to ride (Federal Highway Administration 2010). Unlike BAZ, sharrows are included in the American version of the Manual of Uniform Traffic Control Devices following a recommendation by the National Committee on Uniform Traffic Control Devices to include the treatment (Federal Highway Administration 2010). It was recommended that the centre of the sharrow marking be placed 11ft (3.4m) from the kerb face or the edge of pavement on a street with parallel parking (National Committee on Uniform Traffic Control Devices 2005). The 2009 edition of the MUTCD (American Version) adopted 4 An Evaluation of Bicycle Awareness Zones

21 this recommendation despite only limited research conducted on the 11ft spacing and no research on other spacing options (Federal Highway Administration 2010). 2.3 Video Analysis Video analysis is a common method used to observe and measure the interaction between motorists and cyclists, and the positions they uptake on the road (Alta Planning and Design 2004, Federal Highway Administration 2010 and Houten & Seiderman 2005). A fourth study by Bhat et al. (2009) actually conducted preliminary tests to determine the best method of capturing the lateral position of the motorist and cyclist as they passed a parked motor vehicle. The preliminary testing conducted by Bhat et al. (2009) determined that the most suitable method for collecting video data was a camera mounted on a tripod in a fixed location. The test also found that the video recorders didn t have good depth perception. Therefore, reference markers are required to determine the distance between the cyclist and the parked motor vehicle (Bhat et al. 2009). In the Federal Highway Administration study (2010), calibrations were provided by using either grid lines painted on the street or traffic cones placed on the sidewalk beside the roadway. The sharrows were also used as a calibrator (Federal Highway Administration 2010). In previous video research by Alta Planning and Design (2004), a tripod mounted video was angled at the cyclists. However, they recommended that other studies use an overhead- mounted camera to reduce measurement error (Alta Planning and Design 2004). 2.4 Sharrow s Influence on the Position of Motorists and Cyclists A number of studies demonstrated that sharrows have a positive impact on the position of cyclists and motorists (Alta Planning and Design 2004, Federal Highway Administration 2010 & Houten & Seiderman 2005). Results from the Federal Highway Administration s Evaluation of Shared Lane Markings indicate that cyclists are provided with more operating space in the presence of sharrows (Federal Highway Administration 2010). The increase in operating space has been quantified by Alta Planning and Design (2004). As a result of the markings, the lateral distance between the cyclists and parked cars increased by eight inches. The pavement markings caused an increase of three to four inches in the distance between cyclists and parked cars when a passing vehicle was present. An increase of over two Nicole Eileen Smart 5

22 feet was also achieved between cyclists and passing vehicles (Alta Planning and Design 2004). Houten and Seiderman (2005) also found that cyclists ride further away from parked cars than when no markings were present. 3 BAZ EVALUATION APPROACHES 3.1 Video Analysis Video analysis was used together with crash statistics to test the validity of the TRUM guidelines and to help identify if there are specific safety issues with BAZ. The video analysis was used to establish how the behaviour and interactions compared on the roads with varying traffic volumes and lane widths, and if there were any safety issues. Comparisons were also made between roads to determine if a multilane road had any influence on behaviour, interactions and safety. 3.2 BAZ Questionnaire Survey Motorists and members of bicycle user groups were asked to complete separate questionnaire surveys. The results of the motorist s survey were used to assess the effectiveness of BAZ based on motorist s understanding of BAZ and whether or not they alter their behaviour on a road with BAZ. Cyclists were also tested on their knowledge of BAZ. The survey included a number of questions to establish how safe a cyclist feels in BAZ and any crashes or conflicts they have had on BAZ which could relate back to the safety of BAZ and motorist s behaviour. The questionnaire asked respondents to answer the same questions for bicycle lanes and roads without any bicycle symbols/ lanes, to assess how they compare to BAZ in terms of safety. The guidelines for the provision of BAZ, which are documented in the Traffic and Road Use Management Manual (TRUM), advise that an education campaign should be targeted at motorists and cyclists when BAZ have been installed (Transport and Main Roads 2009). There is no evidence of campaigns in the past. The survey will indicate if a campaign is still required. 6 An Evaluation of Bicycle Awareness Zones

23 3.3 Crash Data Analysis Crash data was analysed to aid in evaluating the safety of BAZ, which until now, has not been the focus of any research. It was expected that defined links would be formulated between crash history, bicycle facility and their locality characteristics. The guidelines in the TRUM (2009) recommend that on roads with an AADT in excess of 3000, a formal bicycle facility be used rather than BAZ. However, a number of local governments are using BAZ on roads with traffic volumes greater than this because of the width and parking constraints on many local roads. The analysis will hopefully determine whether exceeding the thresholds stated in the TRUM, such as installing BAZ on roads with an annual average daily traffic greater than 3000 vehicles per day, has any implications. This project should establish relationships that could be used in TMR s scheduled review of the TRUM operational guidelines for BAZ. 4 VIDEO ANALYSIS 4.1 Methodologies Video analysis was conducted at three Brisbane BAZ sites. The bicycle coordinator from Brisbane City Council was consulted for his advice on which Brisbane BAZ sites may be worth analysing. The coordinator provided a fairly extensive list of sites. The three sites were chosen because they have reasonable cyclist volumes and they have varying traffic volumes (all exceeding the 3000 AADT threshold), parking lane width and parking turnover. One road was multilane. This made the sites suitable for a comparison of the interaction between motorists and cyclists and their positions on the road in different environments. Peak and non peak periods of analysis were compared for each road to assess if there was a change in behaviour. A list of the variables coded for each cyclist at the sites is contained in Appendix A. In summary, the positions of the cyclists relative to passing vehicles, parked cars and the parking edge line were measured. In addition, any conflicts or crashes were described. The coding was largely based on that used by the Federal Highways Administration (2010) and Houten and Nicole Eileen Smart 7

24 Seiderman (2005) who conducted similar studies of an American treatment similar to BAZ, known as sharrows. Three requests for quotes to conduct the video recording were made to traffic data companies. Skyhigh Data were chosen as they have a pro bono agreement for cyclist research and they had appropriate equipment. The video cameras were erected similar to the camera in Figure 1. Figure 1. An example of the video camera set- up. In this study, chalk lines were marked on the road in the parking lane and the traffic lane at 20cm increments measured from the parking edge line. The chalk lines were remarked daily. The video recorded for ten days to ensure enough data. The marking of the road and position of the video camera reflected that used in previous studies by Alta Planning and Design (2004), Bhat et al. (2009), Federal Highway Administration (2010) & Houten & Seiderman (2005) as discussed in section 2.3 of the literature review. The videos were supplied in AVI format and were viewed on a laptop computer. The data for each cyclist was entered into an Excel spreadsheet and was later analysed. 4.2 Characteristics and Locations of Video Sites As previously mentioned, three Brisbane BAZ sites with varying characteristics were analysed. Table 1 compares the sites. 8 An Evaluation of Bicycle Awareness Zones

25 Table 1. Site characteristics at the locations of the video recording Annerley Road Kedron Brook Road Riding Road AADT Note Peak hour traffic (veh/h) Note 2 Lanes/ direction Lane width (m) Parking width (m) Note 1: AADT data is modelling data supplied by TMR Note 2: Peak hour traffic was only counted in the direction of the video recording Annerley Road is a 4 lane arterial road that extends from Ipswich Road, Annerley to Stanley Street, South Brisbane. From the Gladstone Road intersection, Annerley Road has a dedicated bicycle lane to the pedestrian crossing opposite Dutton Park Primary School. At this location the BAZ commences and the BAZ extends to the intersection with Park Road. Further north of Park Road, BAZ is interspersed with sections of bicycle lane. The video site was located in a BAZ section marked on the edge of a parking lane on the northbound side of Annerley Road, opposite Dutton Park State School as seen in Figure 2. The parking in this location appeared to be long term resident parking. a) b) Figure 2. a) Annerley Road Site View b) Location of Video Site on Annerley Road Nicole Eileen Smart 9

26 Kedron Brook Road is a two lane connecting road that links Days Road to Newmarket Road in Wilston. The road passes through a residential precinct, with a few small restaurants to the north of the video site location. BAZ linemarking straddles the parking edge line on both sides of the road. The BAZ section is fairly extensive along Kedron Brook Road, stretching from Main Avenue (430m prior to the video site) to 212m past the video site where Kedron Brook Road meets with Silvester Street. The video site was located on the southbound side of Kedron Brook Road, opposite Jeffery Street. The parking at this site was rarely utilised residential parking. The video site is shown in Figure 3. a) b) Figure 3. a) Kedron Brook Road Site View b) Location of Video Site on Kedron Brook Road Riding Road is a 2 lane local collector road with a few small shops and connects Wynnum Road to Oxford Street/ Lytton Road. BAZ linemarking adjacent to parking lanes extends the entire length of Riding Road. The video site was located on the northbound side of Riding Road, just south of Pashen Street which is on the opposite side of the road as seen in Figure 4. Morningside State School is located off Riding Road near the video site. The site was adjacent to a playing field and a few small businesses were on the southbound side of the 10 An Evaluation of Bicycle Awareness Zones

27 road, opposite the video site. This site had the highest parking turnover of all the sites with most motorists attending the nearby businesses. a) b) Figure 4. a) Riding Road Site View b) Location of Video Site on Riding Road 4.3 Summary of Video Statistics The statistics of the cyclists positions for each analysed road have been averaged and are reflected in Table 2. Nicole Eileen Smart 11

28 Table 2. Summary of Cyclists Positions No. of cyclists in Sample Hrs in Sample Analysis - Cycle Movements With No Interaction with Vehicles Analysis - Cycle Movements Passed by a Vehicle Analysis - Cycle Movements Followed by a Vehicle Site Time Period No. of cyclists no interaction Cyclists' offset to edge line Cyclists' offset to parked car No. of cyclists passed by veh Cyclists' offset to edge line Cyclists' offset to parked car Cyclists' offset to passing veh No. of cyclists followed by veh Cyclists' offset to edge line Cyclists' offset to parked car Annerley Rd Morning Peak Period Annerley Rd Non Peak Period Riding Rd Evening Peak Period Riding Rd 3 days peak and non peak Kedron Brook Rd Morning Peak Period NA NA 171 Kedron Brook Rd Evening Peak Period NA Kedron Brook Rd Non Peak Kedron Brook Rd Non Peak NA 1-40 NA 180 Note: All measurements are mean values and are in centimetres. 12 An Evaluation of Bicycle Awareness Zones

29 4.4 Annerley Road Peak Period For the analysed peak hour period between 7am and 9am on the 27 th April 2011, 89 northbound cyclist movements were recorded on Annerley Road. Cycle movements in the morning peak time are predominantly CBD commuters Annerley Road- Peak Period- Vehicle Interactions During the morning peak period along Annerley Road, there were 52 instances where a vehicle either followed or passed a cyclist. This equates to 58.4% of recorded cyclists interacting with vehicles. The interactions typically involved a car, though there was one incident of a motorbike following a cyclist, one truck followed a cyclist, one truck passed a cyclist and one bus followed a cyclist. The 52 interactions were comprised of 33 vehicles following cyclists and 19 vehicles passing cyclists. For one- third of the cases where vehicles followed cyclists, a queue of traffic did result. This queue ranged from 2 vehicles to 14 vehicles. Table 3 shows the times when the queues occurred. Table 3. Queue lengths and the traffic and cyclist volumes at the times of the queues on Annerley Road. Time (am) No. of Queue Length (number of vehicles) Traffic Volume (veh/ 15mins) Cyclists (Time) Kerb Side Lane Lane 2 6:15-6: (6:22) :30-6: (6:39), 7 (6:41) :45-7: (6:53), 8 (6:52) :00-7: :15-7: (7:42), 5 (7:25), 10 (7:20) :30-7: (7:42), 5 (7:39), 14 (7:39) :45-8: (7:57), 6 (7:50), 6 (7:54) :00-8: (8:14) :15-8: (8:16) Note: The table also contains queues at non- peak time Nicole Eileen Smart 13

30 It can be seen from Table 3 that all queues occurred over a two hour period. The majority of the queues were between 7:15am and 8am where cyclist and traffic volumes were at a peak. For most of the cyclists recorded at this location, no queue formed so the traffic flow was sufficiently dispersed with sufficient overtaking opportunities to avoid queuing. It s not known where the queue started to form as there was a 90m section of BAZ on the approach to the video monitoring point. The mean position of cyclists who were followed was 69cm from the side of the parked car and 49cm from the parking edge line, inside the traffic lane. This offset places the cyclist just within the outer limits of the BAZ symbol. When vehicles were recorded passing the cyclists, the mean position of the cyclist was 60cm from the side of the parked car and 40cm inside the traffic lane. Cyclists seem to be slightly adjusting their lateral position when they sense a car is passing. In passing movements, the mean offset from the passing car to the cyclist was 98cm. Based on these recorded offsets, it appears that even for the constrained parking and travel lane arrangements at Annerley Road, a car could still pass a cyclist without encroaching into the adjacent lane. Notwithstanding these averaged results, about 37% of passing vehicles either drove on or crossed the dividing line. The cyclist s separation between the parked car and passing vehicle can be combined and referred to as the cyclist s envelope, or operating space. Austroads states that the minimum required operating space is 1m, though additional clearance is required to fixed objects and passing vehicles (Austroads 2011). Despite the constrained conditions on Annerley Road, the recorded envelope was 158cm Annerley Road- Peak Period- No vehicle Interactions When the cyclist was not passed or followed by a vehicle, the mean position of the cyclist was calculated at 69cm from the outside of the edge line into the travel lane and 89cm from the side of the parked car. The position of the cyclist in the traffic lane seems to be influenced by the presence of vehicles. The mean offset from the edge line is greatest when there are no vehicles approaching from behind and is least when vehicles are passing. 14 An Evaluation of Bicycle Awareness Zones

31 Consistent with the trend of the offset from the edge line, the mean offset of the cyclist from the parked car increases when there are no vehicle interactions and is smallest when cyclists are being passed. The maximum offset was 89cm and the smallest was 60cm Annerley Road- Peak Period- Conflicts There were 2 conflicts, both of similar nature, which occurred during peak hour. The cyclists rode up the side of the moving traffic which was flowing rather slowly. In both cases, the cyclist allowed 80cm between themselves and the moving traffic. One cycled 40cm from the parked cars, whilst the other cycled 60cm from the parked cars. Both cyclists lost their balance and became unsteady as they passed the traffic. The cyclists were forced to make a fairly abrupt stop to avoid hitting the parked cars Annerley Road- Non Peak Analysis of cyclist movements for Annerley Road in non- peak times was undertaken on Thursday 21st April and Wednesday 27 th April Between 10:30am to 1:00pm on the 21 st April, and 5am to 7am and 9am to 10am on the 27 th April, there were 95 northbound cyclist movements Annerley Road- Non Peak- Vehicle Interactions There was significantly less vehicle interactions during non- peak hour than peak hour, with only 26.3% of cyclists being either passed or closely followed by a vehicle. All interactions involved cars apart from one truck and one bus interaction. 76% of interacting vehicles chose to follow the cyclist, with only six or 24% passing the cyclist. Vehicles are more inclined to follow a cyclist in non- peak hour than peak- hour. There were four instances, or 21% of following cases, which resulted in a queue of traffic forming behind the initial following car. These queues ranged from three vehicles to eight vehicles. Compared to peak hour, there were fewer cases of queues forming and the maximum queue was half the size of the maximum queue during peak hour. As shown in Table 2 the offsets during non- peak conditions were slightly greater than peak time. Nicole Eileen Smart 15

32 The total operating envelope for passing movements during non- peak hour is about 30cm greater than in peak hour. Half of the passing vehicles drove on the centre line and another two crossed the dividing line. This equates to 83% of passing vehicles either driving on or crossing the dividing line, a figure far greater than during peak hour. The two vehicles which crossed the dividing line left 1.4m separation between themselves and the cyclists Annerley Road- Non Peak- No Vehicle Interactions As summarized in Table 2, there is a higher tendency for cyclists to ride further in the traffic lane when there are no vehicles. Offsets for non- peak times with no vehicle interaction were consistently greater than offsets with no vehicle interactions in peak times. No conflicts occurred during the period of analysis. 4.5 Kedron Brook Road Peak Period For the analysed peak hour period between 7am and 9am on the 27 th April 2011, 88 southbound cyclist movements were recorded on Kedron Brook Road. Cycle movements are strongly directed to the CBD with only three southbound cyclist movements recorded in the evening peak period of 3pm to 6pm Kedron Brook Road- Peak Period-Vehicle Interactions There were 27 incidences of vehicle interactions with cyclists during the morning peak hour period of analysis of Kedron Brook Road. This equates to interactions with 31% of observed cyclists. All interactions involved cars, apart from one interaction with a bus and one interaction with a truck. There were no interactions in the evening peak hour period. All interactions were recorded as passing the cyclist. There were no parked cars during the period of analysis and consequently all cyclists rode within the parking lane. As a result, there was no need for a vehicle to follow a cyclist. 16 An Evaluation of Bicycle Awareness Zones

33 The average position of the cyclist when they were passed was 43cm inside the parking edge line. This alignment is on the inner extremity of the 1100mm BAZ symbol. There were no parked vehicles during peak hour so the mean cyclist envelope cannot be calculated. The mean distance between the cyclists and passing vehicles was quite substantial at 171cm. Although the cyclists are well clear of the passing vehicle, motorists are still positioning themselves well within the lane to maximise the offset to the cyclist. Only one vehicle, out of the 27 vehicles that passed cyclists, crossed the centre line of the road. This vehicle allowed 2m separation between the cyclist and themselves and consequently crossed 10cm over the centre line Kedron Brook Road- Peak Period- No Vehicle Interactions Whether or not the cyclist is interacting with a vehicle seems to have little effect on the position of the cyclist. The mean position of the cyclist when there are no interactions was 43cm inside the parking lane versus 50cm inside the parking lane when the cyclists were being passed. Interestingly, a higher percentage of the cyclists who didn t interact with vehicles than cyclists who did interact, rode 1m or more inside the parking lane. There were even four cyclists who rode 1.6m in the parking lane, which is greater than the maximum distance inside the parking lane ridden by a cyclist who was being approached by a vehicle. There were no conflicts observed during the period of analysis Kedron Brook Road- Non Peak A total of eight non- peak hours were reviewed in the 5am to 7am and 9am to 3pm period of analysis on the 27 th April. For two hours, a car was parked in the parking lane and 35 cyclist movements were recorded. For the other hours, only eight cyclist movements were noted and there were no parked cars. Nicole Eileen Smart 17

34 Kedron Brook Road- Non Peak- Vehicle Interactions There were substantially less vehicle interactions during the non peak hour period of analysis than the peak hour period. Only 9% of observed cyclists were involved in interactions with vehicles during the morning non peak- hour period and only one vehicle interaction in the afternoon non- peak hour period. All interactions between vehicles and cyclists during nonpeak periods only involved cars and all interactions were passing manoeuvres (no vehicles were seen to follow a cyclist). The average position of the cyclist when a car passed the cyclist in the morning non- peak hour period was 33cm inside the edge line of the parking lane and 87cm from the edge of the parked car. The average offset between cyclists and passing vehicles for this period was 153cm. During the afternoon period, the cyclists averaged 40cm inside the parking and 180cm offset between cyclists and passing vehicles. The cyclists didn t ride as far in the parking lane as they did during peak hour, however unlike in peak hour, there was a parked car during the morning non peak hour period of analysis. None of the passing vehicles drove on or crossed the centre line of the road, which apart from the one vehicle who crossed the centre line during peak hour, is a very similar outcome to peak hour Kedron Brook Road- Non Peak- No Vehicle Interactions In the morning non- peak hour period, the mean position of cyclists who are being passed is 33cm inside the parking lane which is 20cm further inside the parking lane than cyclists who aren t interacting with a vehicle. At this time, there was a parked car. It seems that the parked car was influencing the cyclist s position as in the afternoon when there was no parked car, the mean position of the cyclist was74cm inside the parking lane. When there are no parked cars, 50% of the cyclists rode between 80cm and 1.2m in the parking lane. When there are parked cars, the furthest a cyclist rode in the parking lane was 60cm. The average offset between the parked car and the cyclist was 87cm when the cyclist was being passed and 107cm when there were no vehicle interactions. Therefore, even with the 18 An Evaluation of Bicycle Awareness Zones

35 wide parking lane, cyclists are still riding in the dooring zone. The dooring zone is the region next to the parked car where conflict could be caused by an open car door, which can be up to 1m wide if the door is fully opened (Austroads 2011). The parked car was 120cm from the parking edge line, so the cyclists had the option to ride out of the dooring zone without encroaching into the traffic lane. 4.6 Riding Road Peak Period Cyclist movements on Riding Road were substantially less than either Annerley Road or Kedron Brook Road. Combining the cyclist movements on Thursday 28 th April and Friday 29 th April, there were only nine cyclists recorded in the 7am to 9am peak hour period and 13 cyclists in the afternoon peak period from 3pm to 6pm Riding Road- Peak Period- Vehicle Interactions In the afternoon peak hour period, 62% of the cyclists were either passed or followed by a vehicle. All interactions involved cars. The interactions were evenly split between passing and following interactions. There was only 2cm difference between the average positions of the cyclists who were followed and the cyclists who were passed. Since the difference is so minimal, and both positions were only slightly in the traffic lane, the cyclists position wouldn t influence the motorists decision to pass or follow. The mean cyclists envelope was 153cm between the parked car and the passing car Riding Road- Peak Period- No Vehicle Interactions The mean positions of the cyclists who were followed and passed, were very close to the parking edge line. However, when there are no interactions the mean position is further in the traffic lane at 32cm, though still within the region of the BAZ symbol. The mean offset of the cyclist from the parked car was greatest when there were no interactions. Even though there were no approaching vehicles, cyclists only allowed an Nicole Eileen Smart 19

36 average of 82cm between themselves and the parked cars and as such, rode within the dooring zone. There was minimal difference in the separation between the cyclist being followed and passed with their offsets being 55cm and 53cm respectively Riding Road- Combined Period Due to the comparatively low numbers of cyclist movements, three days of aggregated cyclist movements were used to obtain better structured representation of cyclist and vehicle positions for Riding Road Riding Road- Combined Period- Vehicle Interactions There were slightly less interactions between cyclists and vehicles during non- peak hour than peak hour with 48% of cyclists interacting with a vehicle. Each interaction was between a cyclist and a car. At 60%, the majority of interactions involved vehicles passing the cyclists rather than following. This is slightly higher than in peak hour where the number of passing and following incidences were even. The average position of the cyclist when being followed was 5cm in the traffic lane, which is no different to peak hour, and following vehicles had sufficient width to pass the cyclist. Despite the positions of the cyclists who were followed being passable, the positions of the cyclists who were passed were more easily passable. The average position was 13cm inside the parking lane which is a 16cm difference in position to that in peak hour. On average, passing vehicles were allowing 127cm separation between themselves and the cyclists. The mean operating space was 183cm between the parked car and passing car Riding Road- Combined Period- No Vehicle Interactions The mean position of the cyclist when there were no vehicle interactions was slightly inside the traffic lane. This is therefore very similar to the mean position of the cyclist when they were followed. 20 An Evaluation of Bicycle Awareness Zones

37 There was little difference in the cyclists mean offset from the parked car when cyclists were either passed, followed or didn t interact with a vehicle. In peak hour, the offset was more varied between interaction types. The mean offset from the parked car was 69cm when there were no interactions, 56cm when cyclists were passed and 67cm when cyclists were followed. When there were no vehicle interactions, the offset was smaller than in peak hour whilst when the cyclist was being passed, the offset was similar to peak hour. Compared to peak hour, the offset when the cyclist was being followed was 12cm larger. Although the sample size for the evening peak period was small, the averaged offsets to the passing vehicle or parked cars were quite similar to those of the larger sample size obtained from three consecutive days of data. This demonstrates there were no cyclists with large offsets, or cars passing with large offsets, in the single day of analysis. 4.7 Validity of TRUM Guidelines Cyclist Envelope Currently Austroads provide little guidance on the clearance requirements of cyclists. Austroads publication, Cycling Aspects of Austroads Guides (2011), suggests that a 1m wide envelope should provide the basis for the design of bicycle facilities as in Figure 5. Figure 5. 1m wide cyclist envelope (Austroads 2011) This 1m envelope takes into consideration the width of the bicycle and allows for tracking. It doesn t allow for the clearance required to parked cars and passing vehicles. Instead, the Nicole Eileen Smart 21

38 guidelines mention that additional clearance is required but doesn t offer any indication of how much is required (Austroads 2011). Figures 6 to 9 have been adapted from Cycling Aspects of Austroads Guides and incorporate the envelopes required on each analysed road which could be used as an indication of the additional required clearance Constrained Situation The results of the video analysis have been used to produce Figure 6 which shows the required offsets in a constrained situation where the parking is 2.5m wide and the traffic lane is 3.3m. From these results, if a cyclist is passing a parked car, a vehicle will need to encroach onto or over the dividing line if it chooses to pass. The majority of motorists will follow a cyclist until there is an overtaking opportunity. The use of BAZ in this situation needs consideration of parking demand, passing opportunities, and traffic volume. Edge of passing car Edge of parked car Peak: 98cm Non Peak: 113cm Peak: 60cm Non Peak: 77cm Figure 6. Cyclist envelope in constrained conditions (adapted from Austroads 2011) Unconstrained Situation (3m parking lane width, 3.3m traffic lane) In an unconstrained situation, the majority of vehicles interacting with a cyclist have sufficient width to pass the cyclist. The resulting envelopes are shown in Figures An Evaluation of Bicycle Awareness Zones

39 Edge of passing car Edge of parked car Peak: 100cm Non Peak: 127cm Peak: 53cm Non Peak: 56cm Note: Offsets in the unconstrained peak condition are based on limited data. Figure 7. Cyclist envelope in unconstrained conditions (adapted from Austroads 2011) Wide Parking Lane (parking lane width: 3.5m, traffic lane: 3.3m) In this arrangement, all interacting vehicles have sufficient width to pass the cyclist. The envelopes for this arrangement are shown in Figure 8 and Figure 9. Edge of passing car Edge of parked car Non Peak: 153cm Non Peak: 87cm Figure 8. Cyclist envelope on a road with a car parked in a wide parking lane Nicole Eileen Smart 23

40 Edge of passing car Parking edge line Peak: 50cm Non Peak: 40cm Peak: 171cm Non Peak: 180cm Note: Offsets in the non- peak condition are based on limited data. Figure 9. Cyclist envelope on a road without a car parked in a wide parking lane (adapted from Austroads 2011) BAZ Symbol (Alignment, Size and Symbol Profile) Table 4. Cyclists offsets from the edge line and kerb on each analysed road Road No. of Cyclists Parked Cars? Mean offset to edge line (cm) Standard deviation of Mean offset to kerb (cm) (-ve values indicate position is left of line) mean offset to edge line (cm) Annerley Road, peak hour 89 Yes Note 1 Annerley Road, nonpeak hour 95 Yes Note 1 Riding Road, peak 13 Yes hour Riding Road, 3 days 73 Yes Kedron Brook Road, 88 No morning peak hour Kedron Brook Road, 3 No An Evaluation of Bicycle Awareness Zones

41 Road No. of Cyclists Parked Cars? Mean offset to edge line (cm) Standard deviation of Mean offset to kerb (cm) (-ve values indicate position is left of line) mean offset to edge line (cm) evening peak hour Kedron Brook Road, 35 Yes non- peak hour Kedron Brook Road, 8 No non- peak hour Mean offset to kerb (cm) 304 Note 1: After passing the parked cars, 95% of cyclists changed their alignment and rode inside the parking lane. Table 4 shows the cyclists mean offset from the edge line for each analysed road and condition. This offset was then represented in terms of the offset from the kerb. If the centre of the BAZ symbol is supposed to represent the intended cyclists line of travel, then it appears the symbol is incorrectly placed since cyclists don t necessarily feel comfortable riding on the edge line. On roads such as Annerley Road with narrow parking (2.5m), cyclists choose to ride to the right of the edge line when passing parked cars. On roads such as Kedron Brook Road with wide parking (3.5m) cyclist choose to ride to the left of the edge line. As noted in Table 4, the cyclists offsets from the kerb are generally around 3m on each road. The overall mean offset from the kerb, calculated using the offset for each road, is 3.04m. This therefore provides argument that the centre of the BAZ symbol should be located 3m from the kerb regardless of the parking lane width. At Annerley Road it was also observed that as soon as the cyclists passed a parked car they changed their alignment to the left of the edge line. It seems reasonable then that for narrow parking lanes of 2.5m or less, the suggested 3 m offset for the symbol should only be used where parking demand is high. With the current placement of the symbol, it is quite likely cars which are parked in a 2.5m wide parking lane will obscure the symbol. This will be rectified by moving the symbol to 3m from the kerb in locations of high parking demand. If there is a deviation from current practices and the bike symbol is not centred on the edge line but could be up to say 0.5m either side of the edge line as for Annerley Road and Kedron Brook Road, then perhaps the spacing of the symbol should be reduced so that the cyclists Nicole Eileen Smart 25

42 have frequent guidance of their alignment. The TRUM currently suggests 200m spacing of bicycle symbols (Transport and Main Roads 2009). A 50m spacing of the symbol may be more reasonable if the centre of the symbol is off the edge line. It may also be worth considering a different stencil which would more clearly define the intended path of the cyclist. Appendix B shows some possible stencils. Each of the stencil options includes some line marking each side of the symbol to give a better guidance for the tracking position of the cyclist. The centre of the current symbol stencil is difficult to judge for cyclists. The commonly used BAZ symbol is 1100mm wide x 1800mm high (Transport and Main Roads 2009). The symbol is 100mm wider than the envelope as shown in Figure 5. If the centre of the symbol is supposed to align with the preferred position of the cyclist, then perhaps consideration should be given to using a wider symbol to take into account more variance in tracking of the cyclists alignment. The standard deviation for the alignment of cyclists varied depending on the traffic conditions and available lane widths but in the most constrained situation on Annerley Road, the standard deviation was about 300mm. On this basis there could be a case for a symbol width of maybe 1700mm. However, as most cyclists and motorists have a poor understanding of the meaning of the symbol, the wider symbol may encourage cyclists to ride on the extremities of the symbol. If this occurs, cyclists will be occupying too much of the traffic lane width. Perhaps consideration should be given to adopting the Transport and Main Roads (2009) standardized 1530mm wide symbol (based on the minimum envelope of 1580mm on Annerley Road in the most constrained condition) with line marking each side of the symbol to give better guidance on tracking. Although there was no video investigation of BAZ on roads without parking lanes, the offset for the constrained conditions are likely to reflect the expected outcome. In Figure 6, the edge line of the parked car would approximate the face of the kerb for roads without parking. Clearly there is not sufficient width to overtake a cyclist with a minimum 1.6m operating envelope unless the lane width is 3.5m or more. The recorded offset data supports the TRUM guidelines for kerb side lanes up to 3.6m wide (refer Appendix C) that the BAZ symbol should be placed more centrally within the lane to discourage overtaking. For lanes of 3.7m width or greater, the symbol placement of 3m from the dividing line provides sufficient offset from the kerb for the cyclist and still sufficient width for motorists to pass. 26 An Evaluation of Bicycle Awareness Zones

43 The standardized 1530mm wide symbol would still be the preference for roads without parking as it better defines the operating envelope than the current 1100mm wide symbol. For wide kerb side lanes, if the wider symbol is to be adopted, a tracking line is suggested to encourage ideal positioning of the cyclist which would allow passing Annual Average Daily Traffic The AADT on Annerley Road far exceeds the 3000 AADT threshold given in the TRUM and was greater than the AADT on the other analysed roads. There were no apparent safety issues on Annerley Road. Only two conflicts were recorded on Annerley Road and the nature of the conflicts could have occurred on lower volume roads. Although there were no apparent safety issues caused by the high volume road, traffic congestion arising from interaction with cyclists was more of an issue on Annerley Road than the other roads. Of the total 89 cyclists in the 2 hour peak period, 52 had interactions with a vehicle (either being passed or closely followed) and in 63% of cases a motorist chose to follow rather than pass the cyclist. The mean offset in peak hour for a passing vehicle was 98cm which is recognised as an appropriate passing distance (Amy Gillett Foundation 2009). As the mean edge line offset of cyclists in a following condition was essentially the same as in a passing condition, clearly the majority of drivers are not confident in passing cyclists in these conditions. One- third of following vehicles on Annerley Road caused a queue, which at times was up to 14 vehicles in length in peak hour. This queue would have formed over a relatively short distance as there was considerable length of dedicated bicycle lane on approach to the 90m monitored BAZ section. On the lower volume Riding Road, queues were infrequent and only two or three vehicles in length. No queues were observed on Annerley Road outside the hours of 6:30am to 8:15am. Annerley Road s high traffic volumes combined with narrow width (parking and adjoining travel lane 5.8m) seem to be the conditions at which the traffic stream could become disrupted by cyclists. However, even with 754 vehicles per hour in the adjoining kerb side lane, the impact at Annerley Road was not significant. Since the BAZ on Annerley Road is interspersed with dedicated bicycle lanes, it is likely that on other roads with similar lane width to Annerley Road, but with longer BAZ sections, the impact on traffic would be greater. The data collected from this survey suggests BAZ are still appropriate for roads with kerb side parking lanes and at least AADT, such as Annerley Road, where BAZ are interspersed Nicole Eileen Smart 27

44 with bicycle lanes. However more research may be needed to confirm traffic impacts on roads with more continuous sections of BAZ, with narrow lanes, and high traffic volume. It is interesting to note that on Kedron Brook Road there were no following cases where the available width was 6.8m, with 955 vehicles per hour during peak hour. The BAZ section preceding the monitoring point was 430m. This indicates that the available width to the dividing line is the main factor that will dictate whether vehicles will pass or follow. Traffic thresholds for the use of BAZ are likely to include parameters such as the available width and for constrained conditions, the length of the BAZ and parking demand Amendments to Parking It was observed in the videos that 95% of cyclists on Annerley Road either rode on the parking edge line or inside the parking lane after passing the parked cars. On this basis, it would be expected that cyclists would move to the left after passing parked cars and allow the following motorist to pass. For sites with a parking lane, high traffic volumes, narrow widths and high cyclist demand, perhaps intermittent no parking locations may need to be considered to allow opportunities for passing. This would produce a similar affect to having BAZ interspersed with dedicated bicycle lanes, as on Annerley Road, where the disruption to traffic was fairly minimal. For sites with wide parking lanes (3m or more), it would be beneficial to encourage motorists to park close to the kerb to maximise the available width to cyclists. Inverted T or L pavement markings could be used to create parking bays 2.1m to 2.3m off the kerb line. The edge line should be retained in these circumstances BAZ through Intersections Where a BAZ is located either side of an intersection, a single continuity line is currently provided for the extension of the parking edge line. A double continuity line with the kerbside line 1200mm off the traffic side line, would guide cyclists to the left of the traffic side line, and increase opportunities for passing. Transitions to the BAZ alignment should be provided each side of the intersection so that the cyclist can reposition themselves. Parking should not be allowed in these transition zones. 28 An Evaluation of Bicycle Awareness Zones

45 4.7.6 Multilane Roads The additional lane on Annerley Road didn t seem to have any impact on the safety of the site. In fact, it may have had a positive impact as it allowed 37% of passing motorists to drive on the dividing line or cross into the adjacent lane in peak conditions. It was surprising that with the high traffic volume on Annerley Road, only 58.4% of cyclists interacted with a vehicle during peak hour. Perhaps some motorists were changing lanes in advance of the cyclists to avoid passing or following the cyclist. The video was not able to capture lane changing more than 30m in advance of the video site. Research has suggested that vehicles increase their travel speed on multilane roads (Rakotonirainy & Schramm 2010). Although speeds were not actually measured, observations indicated that most passing vehicles were travelling only slighter faster than the cyclists in peak conditions. 5 BAZ QUESTIONNAIRE SURVEY 5.1 Methodology The motoring public were approached at parks adjacent to roads with BAZ symbols and asked to participate in the survey. These sites were: Marine Parade, Labrador Eagle Terrace, Shorncliffe The answers provided by the respondents on site were entered into Survey Monkey, a program which assists in developing surveys and analysing the results. A survey specifically for cyclists was ed to organisers of bicycle user groups in the Brisbane and Sunshine Coast Regions, for distribution to their members. Survey Monkey was used for the survey of bicycle user groups. Survey Monkey allowed participants to click on a link to fill out the survey online. Survey Monkey collated the results and presented the results in tables and graphs which were analysed. Nicole Eileen Smart 29

46 The analysis results were compared to the results from a similar survey undertaken by Connell Wagner in 2001, to investigate if there has been a change in the public s understanding of BAZ and the effectiveness of BAZ. Copies of the motorists and cyclists surveys are included in Appendix D and the full results are in Appendix E. 5.2 Survey of Motorists Motorists Survey Sample Size The survey was completed by 70 motorists, 56.5% of whom were also cyclists. However, only 17.1% of these cyclists were frequent cyclists, which have been classified as those cycling in excess of one day a week Motorists Awareness Of the 70 motorists surveyed, only 3 people could not recall seeing the yellow symbol. Neither of these 3 people were cyclists Motorists Understanding Motorists Response Regarding BAZ on a Road without Parking This question allowed respondents to pick one or more possible purposes of the yellow bicycle symbol. As shown by the results in Figure 10, the majority indicated its purpose was to warn of the presence of cyclists but a large proportion incorrectly indicated it denoted a bicycle lane, with 20% of respondents believing that the only purpose was to denote a bicycle lane. Only about 20% recognized the symbol indicated the position cyclists are expected to ride. A small minority, 14.3%, were able to correctly recognize the three purposes indicated in the TRUM, though they all also thought that the symbol indicated a bicycle lane. The correct responses are shown in Figure 10 with ticks. 30 An Evaluation of Bicycle Awareness Zones

47 Figure 10. Responses regarding the purpose of BAZ Motorists Response Regarding BAZ on a Road with Parking The options for the question were the same as for above. This question was to test if respondents viewed the purpose of BAZ differently if there was a parking lane. The difference in responses for BAZ symbols with and without parking are shown in Table 5. Table 5. Variance between responses for symbols with and without parking BAZ on a Road without Parking BAZ on a Road with Parking To warn motorists of the possible presence of 60.9% 49.3% cyclists To denote a bicycle lane 55.1% 33.3% To indicate the position cyclists are expected to ride 23.2% 21.7% To provide continuity in a bicycle route when there 29% 37.7% is not enough space for a full width bicycle lane Other 1.4% 8.7% Nicole Eileen Smart 31

48 Surprisingly, significantly more respondents believe the symbol denotes a bicycle lane on a road without parking than a road with parking. This result is contradictory to what was expected. It was thought that the symbol combined with a parking edge line may lead motorists to believe that it is a bicycle lane. However the edge line doesn t appear to have this effect on motorists Motorists Response Regarding Bicycle Lanes In this question, 90% of respondents correctly answered that the white symbol indicates a bicycle lane. A significant number also responded that the symbol indicated the position cyclists are expected to ride. While this is generally correct, cyclists are only required to use the bicycle lane when it is practical Motorists Behaviour Passing a Cyclist A number of the survey questions tested motorists tendencies in passing a cyclist where there was either a BAZ, bicycle lane, or a street with no bicycle symbols or lanes. As would be broadly expected, motorists indicated that they feel the need to cross the centre line of the road when passing a cyclist riding in a BAZ, or on a street with no bicycle symbols or lanes, more often than a road with a bicycle lane. Motorists seem to recognize that cyclists in a bicycle lane have their own space, unlike a BAZ, and hence there is no warrant for crossing the centre line. However, there were still a high number of respondents who would still cross the centre line to pass a cyclist in a bicycle lane. Motorists seem to react to a road with a BAZ in a very similar way to a street with no bicycle symbols or lanes. A number of respondents commented that they would only cross the centre line of the road if it was legal to do so at that location. Others also stated that the separation they feel necessary to provide to the cyclists depends on the confidence of the rider. Motorists feel far safer passing a cyclist in a bicycle lane when there is oncoming traffic, than a cyclist in a BAZ or on a street with no bicycle symbols or lanes. This mirrors the results of how often a motorist crosses the centre line when passing a cyclist in a BAZ. Again, motorists are recognising the difference in degree of separation between a BAZ and bicycle lane environment. 32 An Evaluation of Bicycle Awareness Zones

49 Motorists feel the least safe passing a cyclist on a road without any bicycle symbols or lanes. Although it is not necessarily desirable to encourage motorists to pass cyclists, BAZ seems to offer a higher level of confidence to motorists Waiting behind a Cyclist Motorists are far less likely to wait behind a cyclist in a bicycle lane than a cyclist in a BAZ or on a road with no bicycle facility. Again it appears evident that motorists are recognising the separation a cyclist has in a bicycle lane, and their behaviour reflects the separation. Motorists seem to react to a cyclist in a BAZ in a similar way to a cyclist riding on a road with no bicycle facilities. The tendency to follow or pass a cyclist was one of the major outcomes tested in the video analysis for different lane widths and the questionnaire results seem to reflect recorded driver behaviour. Refer discussion in section 4.4 to Parking Two questions were directed at testing responses in parking situations: whether or not a motorist checks for cyclists when entering/ exiting parking spaces; and whether a motorist checks for cyclists before opening their car door. In general, motorists seem to be aware of the risk to cyclists in parking situations. Irrespective of the type of bicycle facility, the majority of respondents (64%) indicated they check for cyclists all the time when they entered or exited a car parking space. Slightly fewer respondents (54%) checked for cyclists all the time, regardless of the facility, when opening a car door Motorists Speed This question tested the motorists perceived need to slow down when passing cyclists either in a BAZ, bicycle lane, or a street without bicycle symbols or lanes. Responses indicate that the separation provided by a bicycle lane has an effect on the motorist s perceived need to reduce their speed. Motorists reduce their speed more often on roads with BAZ and no bicycle facilities than on roads with a bicycle lane. The number of respondents for each frequency was pretty similar for BAZ and roads with no bicycle facilities. Only 30% of respondents answered they reduce their speed all the time regardless of the facility Motorist s Opinion of the Effectiveness of BAZ In this question, there was an overwhelmingly positive perception of BAZ. 88% of respondents either strongly agree, agree or somewhat agree that BAZ make a safer road Nicole Eileen Smart 33

50 environment for cyclists than a road without any bicycle symbols. The most common agreement was strongly agree which was answered by 46% of the respondents. These results are not entirely reflective of those obtained in other questions. The safety would largely be influenced by motorists caution when entering/ exiting a parking space or opening their car door in a parking situation, and their speed around cyclists. In the previous questions, motorists answered that they behave very similarly regardless of whether there is a BAZ or a road with no bicycle symbols. On this basis, they have not provided evidence to support the statement, Bicycle Awareness Zones make a safer road environment for cyclists than a road without any bicycle symbols at all. It is assumed that when motorists first see the statement, it seems obvious that a bicycle facility would make a safer road environment than one without Motorists Response to a Scenario The scenario presented in the question involved rating responses to the changes in parking or bicycle facilities that a road authority should implement if a cyclist had been sideswiped whilst passing a parked car. The scenario demonstrates the value of parking to the public. The least preferred response was removing parking to enable the installation of bicycle lanes. The most preferred treatment was split equally between widening the street to retain parking and enable bicycle lanes; and retaining all parking and installing a bicycle awareness zone. The second and third preferences for the two aforementioned options were also very balanced. Assigning points to the degree of preferences found that bicycle awareness zones were most preferred. However there was not a clear preference, and as such, it is only recommended that parking not be removed to install a bicycle lane Motorists Comments with Respect to BAZ Respondents were encouraged to comment on positive aspects or any problems or issues with BAZ. The full range of comments is included in Appendix F. The overall consensus of motorists was cyclists should not be allowed to ride on the road if there is not a bicycle lane. A number of people with this view also made mention that cyclists shouldn t have the right to cycle on the road since they don t pay registration. 34 An Evaluation of Bicycle Awareness Zones

51 One motorist made a very interesting point with regards to the colouring of BAZ. They thought it was inconsistent to paint an area to be used by cyclists in yellow, since the no standing line is also painted yellow. 5.3 Survey of Cyclists Cyclists Survey Sample Size The survey of cyclists was completed by 311 cyclists who are members of Brisbane and Sunshine Coast bicycle user groups Cyclist s Awareness A high proportion of survey respondents, 95.8%, have cycled along roads with BAZ. The remaining respondents have either not ridden, or couldn t recall if they had ridden along a BAZ section Cyclists Understanding Cyclists Response Regarding BAZ There seems to be a fairly strong understanding of BAZ amongst the cycling fraternity. 84.4% of respondents are aware that BAZ don t provide an area exclusively for cyclists and 88.8% realise that they are not legally required to ride in a BAZ. The question posed to motorists about their understanding of the purpose of BAZ was also asked of cyclists. 29.4% of survey respondents correctly selected all 3 purposes of BAZ. Respondents were most aware that the symbols warn motorists of the possible presence of cyclists. Of the correct answers, the fewest respondents, 55.4%, knew that the symbols indicated the position cyclists were expected to ride. It was surprising that 30.9% incorrectly thought that BAZ denotes a dedicated area for cyclists. This is inconsistent with the response to the earlier question where most respondents were aware that BAZ don t provide an area exclusively for cyclists Cyclists Response Regarding Bicycle Lanes It appears that most cyclists understand the difference in exclusivity between BAZ and bicycle lanes, with the cyclists recognising that BAZ don t provide an area exclusively for Nicole Eileen Smart 35

52 cyclists, and also answering that bicycle lanes do provide an area exclusively for cyclists. However, only 46.84% knew that bicycle riders are legally required to ride in bicycle lanes. However, it is acknowledged that there may be some degree of confusion with this question since bicycle riders are legally only required to ride in bicycle lanes when it is practical Cyclists Response Regarding the Difference between BAZ and Bicycle Lanes Cyclists have a very negative opinion of the motorist s understanding of BAZ and bicycle lane linemarking. 81.4% of respondents disagree, to various extents, that the average motorist understands the differences between linemarking for bicycle lanes and bicycle awareness zones. The highest response was cyclists strongly disagree that motorists understand the difference in linemarking. These results are supported by the responses given by motorists, where it was a common response that the yellow BAZ symbol indicates a bicycle lane Cyclists Response Regarding the Safety of Bicycle Facilities There isn t an overwhelmingly general perception of safety amongst cyclists who cycle along a main street with no bicycle facilities/ symbols. One- third of respondents only feel safe some of the time or occasionally and almost 40% rarely feel safe or never feel safe. In contrast, and as would be expected, cyclists generally feel safe on a quiet street with no bicycle facilities/ symbols, with 85.5% of respondents indicating a positive degree of safety. Only 1.5% feel safe rarely or never on a quiet street with no facilities. Half of the respondents answered positively about their degree of safety in a BAZ, with the highest response being the cyclists feel safe much of the time. 11.9% of respondents rarely or never feel safer riding in a BAZ. Cyclists feel most safe in bicycle lanes, with just over three quarters of respondents answering they feel safe all the time, most of the time (the highest response) or much of the time when riding in a bicycle lane. 2.6% of respondents rarely or never feel safe in a bicycle lane. Comparing the percentage of respondents answering positively and negatively about their perception of safety, a quiet street with no bicycle facilities/ symbols was deemed to be most safe, closely followed by bicycle lanes. Since there was a fairly high proportion of cyclists who rarely or never feel safe on main thoroughfares with no bicycle facilities/ symbols, this reinforces the need to provide a bicycle lane or BAZ, both of which have been perceived to be safer in this survey. 36 An Evaluation of Bicycle Awareness Zones

53 In summary, on a main thoroughfare, cyclists feel most safe in a bicycle lane. On a quiet street, cyclists don t seem to require a facility in the form of a BAZ or a bicycle lane to feel safe. Where possible on a non- quiet street, a bicycle lane should be installed instead of a BAZ. Cyclists level of safety therefore mirrors the guidelines in the TRUM which states that BAZ must only be used if bicycle lanes cannot be provided due to constrained road space (Transport and Main Roads 2009). Cyclists were asked to agree or disagree with the statement, Bicycle Awareness Zones make a safer road environment for cyclists than a road without any bicycle symbols at all. Respondents generally believe that BAZ do make a safer road environment, with 64.7% of respondents strongly agreeing, agreeing or somewhat agreeing with the statement. The most common agreement was somewhat agree Cyclists Response to Cycling Hazards Respondents were asked to comment on the hazards they face when cycling. Over half of the respondents took this opportunity to voice their concerns. Responses are outlined in Appendix G. Some of the identified hazards are associated with BAZ and a discussion of these hazards is presented below. A high proportion of cyclists are concerned that motorists are not aware of the presence of cyclists and regularly turn left in front of a cyclist, thereby cutting off the cyclist. Although this issue is largely the motorist s responsibility, rather than a design deficiency, BAZ symbols should be placed more frequently, and particularly in the lead up to an intersection or frequently used driveways such as takeaway shops. Many cyclists believe that motorists don t know the road rules and in particular the cyclist s right to cycle on the road and the requirement to give way to cyclists. They also believe that motorists need to be educated about the spacing they should be providing between the cyclist and their vehicle. One respondent suggested that there should be an additional section in the licensing test to ensure motorists know how to drive around vulnerable road users such as cyclists and pedestrians. Perhaps after education, there will be a reduction in unruly behaviour directed to cyclists from motorists, which is currently a common hazard experienced by cyclists. Nicole Eileen Smart 37

54 Cyclists feel at risk when their lane ends at hazardous locations such as at intersections. For this reason, further investigation needs to be conducted to develop a consistent approach for providing cyclist facilities through intersections. This should also include roundabouts which were another hazard identified by cyclists Cyclists Response Regarding Conflicts Cyclists were asked to comment on how often they have been involved in conflicts on roads with no bicycle facilities, roads with BAZ, and roads with bicycle lanes. Conflicts were defined as being a sudden change in speed or direction to avoid the other party. Weighing up the number of regular conflicts (daily and weekly conflicts) to infrequent conflicts (conflicts occurring rarely or never), bicycle lanes are safer than BAZ. A quiet street with no bicycle facilities/ symbols is the safest, and a main thoroughfare with no bicycle facilities/ symbols is the least safe. Additional comments provided by the respondents about the conflicts they have been involved in are presented in Appendix H Cyclists Response Regarding Crashes Cyclists were also asked to comment on how often they have been involved in crashes on the various road types and markings. The results indicated that crashes were three times more common on a main thoroughfare with no symbols than the other road types, which again reinforces the inadequate safety on these types of roads. In terms of crashes, a quiet street with no bicycle facilities/ symbols is as safe as a road with a bicycle lane and a road with a bicycle awareness zone. Additional comments provided by the respondents about the crashes they have been involved in are presented in Appendix I Cyclists Response Regarding Motorists Behaviour in BAZ Cyclists were asked what ways BAZ affects motorists behaviour compared to riding where there is no bicycle facility/ symbol. The general consensus amongst cyclists is that BAZ have only a minimal affect on motorist s behaviour. Figure 11 shows the behavioural changes cyclists were asked to assess and the results. 38 An Evaluation of Bicycle Awareness Zones

55 Figure 11. The effect of BAZ on motorists behaviour Respondents were also asked to list any other changes in motorist s behaviour on roads with BAZ compared to a road without any bicycle facilities/ symbols. A number of cyclists commented that the BAZ symbol seems to legitimise the cyclist s right to cycle on the road and as a result, they have less abuse directed at them Cyclists Response Regarding Motorists Behaviour in Bicycle Lanes Cyclists were also asked if they thought bicycle lanes affect motorist s behaviour compared to roads with no facilities/ symbols. As with BAZ, bicycle lanes are perceived to have little effect on behaviour with the highest responses being a combination of rarely and not at all. The only behaviour that was made more frequently (combination of all the time, most of the time and much of the time) than infrequently (combination of rarely and not at all) was motorists giving cyclists more space when passing cyclists. However, this change could be due to motorists being guided to give the cyclist more space by the bicycle lane edge line, rather than the motorist making a conscious decision to give more space to the cyclist. Again some respondents commented that motorists seem to be more patient and courteous as the cyclist s right to cycle on the road is evident Cyclists Response Regarding Issues with BAZ When given the opportunity to comment, many cyclists mentioned that motorists believe that the cyclists should stay to the left of the parking edge line if there is a line. Consequently, Nicole Eileen Smart 39

56 when a cyclist is forced into the traffic lane to avoid a parked car or debris, motorists have little respect for them. According to cyclists, motorists need to be better informed about BAZ; how to drive in the presence of cyclists; and be made aware that cyclists may need to change their position to avoid hazards Recommendations Made by Cyclists move the BAZ symbol to the middle of the traffic lane so that motorists are more likely to see it; include the words LOOK or BEWARE on the pavement at the start of the section with BAZ; at the lead up to roundabouts, paint watch for cyclists on the pavement; remove Raised Retroreflective Raised Pavement Markers from the left of the BAZ symbol as these pose a hazard to cyclists. 5.4 Comparison between Motorists and Cyclists Responses The cyclists comments and responses to survey questions, such as the question asking the cyclists to agree/ disagree with the statement that the average motorist understands the differences between linemarking for bicycle lanes and BAZ, of which the majority of cyclists disagreed to a certain extent, are supported by the motorists lack of understanding of BAZ as evident in their survey responses. Although most respondents in both the motorists and cyclists surveys agree, to a certain extent, that BAZ make a safer road environment than a road without any bicycle symbols at all, motorists were far more in agreement. 88% of motorists either strongly agree, agree or somewhat agree, with the majority strongly agreeing, with the statement, whilst only 64.7% of cyclists agree to an extent, with the majority only somewhat agreeing with the statement. 5.5 Comparison with Connell Wagner Study Following the analysis of the motorists responses to this survey, some interesting observations are made when compared to the survey conducted by Connell Wagner in In 2001, only 55% of respondents had observed BAZ symbols compared to 96% in the 2011 survey. This result reflects the increased implementation of BAZ since An Evaluation of Bicycle Awareness Zones

57 The Connell Wagner study found yellow symbols lead to positive driving changes in 82% of drivers. Motorists stated that they watched for cyclists, drove more cautiously and moved away from the edge line in the presence of BAZ. The IAP project s study found the behaviour of drivers in BAZ locations is positive, though the responses indicated that the symbol didn t really have any influence on behaviour as the motorists are behaving similarly on roads without any bicycle symbols. In the Connell Wagner questionnaire, all cyclist respondents believed that the symbol was effective in changing behaviour. This perception seems to have changed as cyclists aren t noticing any changes in motorist s behaviour between BAZ locations and roads without any bicycle symbols or lanes. It was however commented in both surveys that the BAZ symbols endorse cyclists legitimate right to be on the road. 5.6 Conclusions of Questionnaire Survey Although awareness of BAZ has greatly increased, and understanding has improved since the Connell Wagner study, the surveys highlighted the need for an education campaign. Although cyclists had a significantly better understanding of bicycle facilities, an education campaign also needs to be directed at cyclists. The education campaign should explain the purposes of the various bicycle facilities; how motorists should drive in the presence of a cyclist in the various facilities; inform motorists of the reasons cyclists may need to cycle further in the traffic lane; and also advise cyclists of the ideal position they should be riding in a BAZ. Once motorists and cyclists are better informed, the effectiveness of BAZ will be improved. As would be expected, cyclists feel safer riding in a bicycle lane than a BAZ. However, BAZ were still considered to be safe by cyclists. The number of conflicts had by cyclists in BAZ and bicycle lanes suggested that bicycle lanes were safer than BAZ. In terms of crashes, the safety of BAZ and bicycle lanes was similar. BAZ don t seem to be affecting the motorists behaviour. Many motorists behave similarly on roads with BAZ and roads without any bicycle lanes/ symbols. The motorists responses suggested that their behaviour is generally safe, with a high proportion of respondents reducing their speed around cyclists, and checking for cyclists when entering/ exiting parking and as they open their cars doors. The results of the cyclist survey supported this finding with the majority suggesting that there is no change in motorists behaviour on roads with BAZ Nicole Eileen Smart 41

58 compared to a road without any bicycle symbols/ lanes. The cyclists survey didn t ask cyclists to specifically rate the motorists behaviour, so the results of the motorists behaviour in the motorists survey can t be validated. Despite the responses to a number of survey questions indicating that BAZ are no safer than a road without any bicycle symbols/ lanes, both motorists and cyclists agreed that BAZ makes a safer road environment for cyclists, than a road without any bicycle symbols at all. In summary, the results indicated that a BAZ or bicycle lane should be provided on a main thoroughfare, with a bicycle lane being the first preference. A quiet street however, does not seem to require a bicycle facility and installing a BAZ would unlikely alter motorist s behaviour. 6 CRASH DATA ANALYSIS 6.1 Methodology The crash data for all cyclist crashes occurring in Queensland between 2004 and 2009 were obtained using WebCrash, an online database containing crash information for crashes which occurred on Queensland s roads. The crash data didn t indicate the facility at the crash location. The 4000 Queensland cyclist crashes were plotted on MapInfo but only the crashes occurring in South East Queensland were selected, since BAZ are predominately used only in South East Queensland. This reduced the number of crashes to approximately Bicycle facility maps were requested from Brisbane City Council, Gold Coast City Council and Sunshine Coast Regional Council. Gold Coast City Council doesn t have such maps and Sunshine Coast Regional Council was unable to provide the maps within the timeframe assigned for the analysis. As a result, only the 1100 or so crashes occurring on Brisbane City Council roads were analysed. MapInfo was used to tabulate the crash and the associated facility on the road where the crash occurred. The MapInfo table was used to determine the number of crashes on BAZ, bicycle 42 An Evaluation of Bicycle Awareness Zones

59 lanes, or roads where there is no bicycle facility. The analysis mainly focussed on the 100 crashes which occurred on either a BAZ or bicycle lane section. Network traffic modelling data was used to determine the AADT at each crash location. Each crash site on a road with a bicycle lane or BAZ was located on aerial imagery using MapInfo and Google Maps. Site characteristics were recorded for each crash site, such as whether or not the crash occurred at mid- block or an intersection; whether or not there was parking at the crash location; an estimation of the parking turnover based on the land use in the vicinity of the crash; whether there was a parking edge line; parking and traffic lane widths; and the gradient of the road. All mid-block crashes were split into BAZ or bicycle lane sites, and similarly for the intersection crashes. The crash report for each mid-block crash was reviewed to assist in indentifying the contributing factors in the crash. The abovementioned site characteristics of each mid-block crash were compared, to determine if there were any links between certain site characteristics and crashes. The crash reports were not reviewed for any intersection where only a single crash occurred, due to the large proportion of intersection crashes and the fact that intersections are already known to be higher risk conflict zones (Forester 1994, p 97). Any trends in BAZ crashes were compared to trends in bicycle lane crashes. Single crashes which occurred on roads with no bicycle treatments were not analysed due to the large number of crashes. Mid-block and intersection locations with multiple crashes were tabulated for each facility. Crash sites on roads without a bicycle facility were included in this section of the analysis. The crash reports for each crash occurring at a single location were compared. Similar reasons for the crash could indicate a site specific problem, rather than random driver error. Sites with multiple crashes were compared for each facility to indentify if there were certain site characteristics influencing a crash for a particular facility. This would complement the analysis of sites with a single crash. A summary of the crash statistics for the Brisbane City Council region is shown in Table 6. Nicole Eileen Smart 43

60 Table 6. Brisbane City Council crash statistics Midblock Intersection Total Crashes Single Crash Sites Multiple Crash Sites Single Crash Sites Multiple Crash Sites No facility Not analysed 2 Not analysed Bicycle lane BAZ Note: Single incident crashes where there was no bicycle facility (BAZ or bicycle lane) were not analysed. 6.2 Crash Rate Calculations To enable comparison between the number of crashes occurring on BAZ and bicycle lane sections, crash rates have been calculated in terms of the number of crashes per kilometre per year of the respective facility type and are shown in Table 7. Table 7. BAZ and bicycle lane crash rates Facility Crashes per km of facility Crashes per km per year BAZ 0.58 (70.5km of BAZ) 0.12 Bicycle Lanes 0.67 (83.1km of bicycle lanes) 0.13 The above calculations suggest that cycling in a BAZ is no more hazardous than cycling in a dedicated bicycle lane. The number of crashes per kilometre per year of BAZ and bicycle lane is very similar. However, it is acknowledged that the calculations don t take into account the number of cyclists or vehicles on that section of road. The number of cyclists and vehicles could have a fairly significant impact on the comparison, since it is expected that the number of crashes would be proportional to the usage of the route. A more detailed assessment of the crash types as outlined below, suggests that some crashes would have been preventable if a bicycle lane was used in place of a BAZ. 44 An Evaluation of Bicycle Awareness Zones

61 6.3 Analysis of Mid- Block Crashes General Analysis of Mid- Block Crashes Table 8. Proportion of total crashes occurring on mid-block sections Facility % of Total Crashes Occurring on Mid- Block BAZ 30% of total BAZ crashes Bicycle Lanes 10.7% of total bicycle lane crashes Again, with only a small dataset, it is difficult to make any defensible conclusions, though the data is indicating that mid-block crashes are more likely to occur on BAZ as seen in Table 8. There was a higher number of dooring incidents on BAZ mid-block sections and this would be influencing the outcome. Typically, bicycle lanes are only installed adjacent to parking where a safety zone of 0.4m to 1m is available next to the parking (Austroads 2011). Therefore, it is logical that there should be less dooring incidents associated with a bicycle lane Detailed Assessment of Mid-Block Crashes The specific details of each mid-block crash in either a BAZ or in a bicycle lane has been undertaken to check for trends or common features. Table 9 outlines the crashes which occurred on mid-block sections of BAZ. Refer Appendix J for DCA coding. Table 9. Crashes which occurred on mid-block sections of BAZ DCA Number Crash Description Preventable? Code of Crashes Car reversed out of driveway and hit a cyclist This could have occurred even if there was a bicycle lane Driver opened car door into the path of the cyclist and the cyclist hit the door Bicycle lane could help to avoid this situation Driver opened car door into the path of the cyclist and the cyclist hit the door Bicycle lane could help to avoid this situation Driver opened car door into the path of the cyclist Bicycle lane could help to avoid this Nicole Eileen Smart 45

62 DCA Number Crash Description Preventable? Code of Crashes and the cyclist hit the door situation. Driver opened car door into the path of the cyclist and the cyclist hit the door Bicycle lane could help to avoid this situation Cyclist left the footpath to ride in the BAZ, and in doing so, collided with a car in the traffic stream Since the facility was a BAZ, the cyclist entered into the traffic stream. A bicycle lane would have provided some separation for the cyclist from the traffic stream A car made a u- turn in front of a cyclist. The cyclist swerved around the car and crashed Unknown if a bicycle lane could have prevented this crash Cyclist swerved around a stormwater drain grate and rode into the side of a bus A bike lane would have given the cyclist more clearance between the traffic. All grates adjoining BAZ or bike lanes should be bike friendly A car turned into a driveway in front of a cyclist who was travelling in the same direction and This is likely to occur even in the presence of a bicycle lane. clipped the cyclist A car turned into a driveway in front of a cyclist who was travelling in the same direction and This is likely to occur even in the presence of a bicycle lane. clipped the cyclist A cyclist was sideswiped by a car Less likely to occur in the presence of a bicycle lane Driver opened car door into the path of the cyclist and the cyclist hit the door Bicycle lane could help to avoid this situation. Of the 12 mid- block crashes that were associated with a BAZ, it is possible that eight of these crashes could have been avoided if there had been sufficient width to provide a dedicated bicycle lane. Some of the accidents occurring on BAZ have been assigned DCA codes which appear to be inconsistent with the crash description. It appears that although there were only two accidents coded as being dooring crashes (DCA code 604), there were actually five dooring incidents. As a result, dooring was the most common crash type along mid- block sections of BAZ. 46 An Evaluation of Bicycle Awareness Zones

63 DCA codes 309 and 408 only occurred at BAZ sections. The crashes coded as DCA 309 are unlikely to be solely related to a BAZ. Rather, the cause of this crash is more attributed to a lack of attention by the driver. Perhaps installing the BAZ symbol at more frequent intervals would remind drivers of the presence of cyclists. In the case of the crash coded as DCA 408, a bicycle lane may have helped to prevent this crash as described in Table 9. The crashes which occurred on mid- block sections of a bicycle lane are presented in Table 10. Table 10. Crashes which occurred on mid-block sections of bicycle lane DCA Number Crash Description Code of Crashes A car swerved into the bicycle lane and sideswiped the cyclist travelling in the bicycle lane Driver opened car door into the path of the cyclist, though in this case the driver was illegally parked A car pulled out of parallel parking and into the path of the cyclist A cyclist was riding in the bicycle lane against the traffic. A car pulled out of a driveway and only looked at oncoming traffic and subsequently the car hit the cyclist A car moved left into a taxi rank and cut off the cyclist who was in the bicycle lane A cyclist was riding in the bicycle lane against the traffic. A car pulled out of a driveway and only looked at the oncoming traffic and subsequently the car hit the cyclist In the case of the bicycle lane crashes, excluding the cases where the cyclists were moving against the traffic, driver negligence was the cause of the crash BAZ Dooring Incidents Although there was a parking edge line at all crash locations, the width of the traffic lane was minimum in all cases, and it is likely the cyclists rode to the left of the edge line rather than on the edge line, which makes them more susceptible to dooring. Based on the crash data, it appears that the combination of minimum parking and traffic lane widths seems to be associated with dooring incidents, as this was a common factor in all dooring incidents which occurred on BAZ. Nicole Eileen Smart 47

64 It is not possible to draw any conclusions about the effect of parking turnover in dooring incidents. The project timeframe didn t allow site visits to the crash sites to assess the parking turnover, and so only aerial imagery was used to gain a vague indication of the parking turnover. Based on the aerial imagery, it appears that 3 of the 5 dooring incidents occurred at sites with fairly continuous parking demand AADT at Mid-Block Crash Locations Table 11. Mid-block crash data relationships with traffic volumes < 3000AADT AADT > AADT BAZ crashes 9 (7P, 2NP) 3 (3P) 0 Bicycle lane crashes 3 (1P, 2NP) 1 (1P) 3 (3NP) P- Parking NP- no parking As seen in Table 11, of the 12 crashes which occurred on mid-block sections of BAZ, 9 occurred at sites where the AADT is less than the TRUM threshold of As there are no trends to suggest a higher crash history on higher volume roads, there doesn t appear to be sufficient justification to apply an AADT threshold for the use of BAZ. The 3000 limit currently in the TRUM appears to be invalid. Brisbane City Council use BAZ on a number of four lane roads with volumes in excess of AADT. There is no data to indicate that BAZ are unsuited on these high volume roads. Under Austroads Guide to Traffic Engineering Part 14: Bicycles, dedicated bike facilities are generally not provided on roads where the AADT is less than 3000, presumably because the risk to cyclists is deemed to be low (Austroads 1999). So it is surprising that a large proportion of the crashes occurred on roads that have less than 3000 AADT. Perhaps the threshold for not providing facilities should be less than 3000 AADT Multiple Crashes at Mid-Block Locations There were no mid-block bicycle lane crash sites with multiple crashes. There was one case of multiple crashes at a mid- block section of a BAZ and two sites with multiple crashes on roads without any bicycle facilities. At each site with multiple crashes, the crashes occurred 48 An Evaluation of Bicycle Awareness Zones

65 due to varying reasons which indicates that there wasn t a site specific issue contributing to the crash. 6.4 Analysis of Intersection Crashes General Analysis of Intersection Crashes Table 12. Proportion of total crashes occurring at intersections Facility % of Total Crashes Occurring at Intersections BAZ 70% of total BAZ crashes Bicycle Lanes 89.3% of total bicycle lane crashes The values in Table 12 suggest BAZ through intersections are no more hazardous than a dedicated bicycle lane. Irrespective of whether there is a BAZ or bicycle lane at intersections, crashes are significantly more likely to occur at intersections rather than mid- block. The high incidence of crashes at intersections is supported in research by Forester (1994) who suggests that 95% of car-bike collisions occur as the result of turning or crossing movements, which are generally located at intersections Multiple Crashes at Intersections There were only two multiple crash sites at intersections with BAZ and the crashes at one of the sites were unique and not due to a site specific issue. There were a high number of multiple crashes at intersections with bicycle lanes and at intersections with no bicycle facilities. Of the multiple crashes at intersections with bicycle lanes, six crash sites may have contributed to the crash, whilst at the multiple crash sites at intersections on roads without any bicycle facilities, site specific issues at four of the sites may have contributed. Of the 11 multiple crash sites at intersections, five involved a cyclist travelling downhill on approach to the intersection and vehicles pulling out in front of the cyclist or turning across the path of the cyclist. In these cases, the grade ranged from 2.8% to 12%. Crashes of this nature occurred on roads with either BAZ, bicycle lanes or no bicycle facilities. Four sites seemed to involve motorists not observing and giving way to cyclists who were on the far side of two traffic lanes. The other apparent site specific issues involved: Nicole Eileen Smart 49

66 motorists turning left and cutting off the cyclist who was on the motorist s left hand side; motorists turning left from a left turning lane which is adjacent to a right turning lane and failing to see the cyclist approaching from the motorists right hand side; motorists failing to give way to cyclists on a roundabout that is particularly small and doesn t satisfy the minimum required dimensions for the central island diameter (10m) and circulating carriageway width (7.6m) as stated in the Road Planning and Design Manual: Chapter 14 Roundabouts (2006); motorists turning left from a slip lane and failing to give way to a cyclist. Some sites appeared to have multiple site specific issues. The crash locations and a detailed discussion of the crashes at each site, are presented in Appendix K Roundabouts Of the intersection crash locations where multiple crashes occurred, 32% of the locations were at roundabouts. However, only 10% of all bicycle lane and BAZ crashes at intersections were at roundabouts. Therefore, for crash locations where multiple crashes occurred, roundabout accidents are statistically overrepresented, since the number of roundabouts is comparatively small compared to all other intersection types. It should be noted that at each roundabout where a crash occurred, there was no cyclist linemarking in the roundabout. 6.5 Conclusions of Crash Data Analysis Few conclusions have resulted from this research, and any conclusions drawn have a limited degree of confidence. Primarily, this has been due to a lack of available data. The numbers of crashes per kilometre per year for BAZ and bicycle lanes were found to be very similar. As such, there is no evidence of a serious safety concern with BAZ. To gain a true understanding of the safety of each facility, ideally the result should be compared to the number of motorists and cyclists using each BAZ and bicycle lane route. Requests were made for data which would allow such a comparison, however the data was not available. The available crash data indicates that mid-block crashes are more likely to occur on a BAZ section rather than a bicycle lane section. BAZ sections had a significantly higher proportion 50 An Evaluation of Bicycle Awareness Zones

67 of dooring incidents than bicycle lanes. This would be influencing the higher percentage of mid- block crashes on BAZ than bicycle lanes. All dooring incidents occurred on roads where there was a combination of minimum parking and traffic lane width. Since dooring incidents could be prevented with the use of a bicycle lane, it is implied that bicycle lanes are safer than BAZ over mid- block sections. The high proportion of crashes at intersections is reflective of previous research. Since intersection crashes are significant for BAZ and bicycle lane approaches, this doesn t indicate a particular safety issue with respect to BAZ. The crash analysis indicates that there are no trends suggesting that BAZ crashes are more likely to occur on roads with a higher AADT than the current TRUM threshold of As such, the TRUM threshold seems to be invalid. Multiple crashes at mid-block sections where there was either a BAZ, bicycle lane or no bicycle facility were compared. Multiple crashes are more likely to indicate a site specific problem, rather than random driver error. However, it was found that many crashes were actually random and couldn t be linked to any site characteristics. The analysis did not indicate that one particular facility was affected by certain characteristics more than another facility. It was found that cyclists seem to be particularly at risk when they are travelling downhill on approach to an intersection. Crashes of this nature were not associated with one facility; rather they occurred on BAZ, bicycle lanes and roads with no bicycle facilities. Of the 11 locations which seem to have a site-specific issue at an intersection, the downhill grade appeared to be a factor in five crashes. The grades at these crash locations ranged from 2.8% to 12%. In each case, the motorist was either turning right across the path of the cyclist, or left in front of the cyclist, and failed to give way to the cyclist. It is assumed that the motorists are either not seeing the cyclist due to the pace they are travelling or they are misjudging the speed of the cyclists. Another common crash on all facility types involved a motorist failing to give way to a cyclist as the motorist turned right across two lanes of traffic. It seems that the motorist s visibility is Nicole Eileen Smart 51

68 restricted perhaps due to other cars blocking the motorist s line of sight to the cyclist. This type of crash occurred in three out of 11 crashes. Perhaps there is a case for a media campaign that could highlight visibility issues with two objectives; the first being for cyclists to be more conscious of the need for high visibility clothing/ lighting; and for motorists to be more cautious and aware of the possibility of cyclists in the area Data Collection by Road Authorities This research, and the findings, could be significantly improved if there was more available data. Local governments would play an important role in the supply of the data. It is suggested that a register is maintained for the location and date of each BAZ or bicycle lane installation. This would not only assist with the research of bicycle facilities but also help local governments review and maintain the facilities. Cycle numbers and the AADT should be recorded for at least the most important cycle routes. Again, this would be important from a research perspective but local governments could also use the counts to monitor the popularity of routes which may help in prioritising development of the routes. 7 CONCLUSIONS 7.1 General As part of a review of the TRUM, three project elements were used to test the relative safety and effectiveness of BAZ. Project elements consisted of video analysis of three BAZ sites; a questionnaire survey of motorists and cyclists about their knowledge of and experiences with BAZ and other bicycle facilities; and crash data analysis. 7.2 Video Analysis Video analysis was conducted at three Brisbane BAZ sites; Annerley Road, Riding Road and Kedron Brook Road. Video cameras recorded the interactions between motorists and cyclists and their positions on the road. The sites chosen for analysis had varying traffic volumes all 52 An Evaluation of Bicycle Awareness Zones

69 exceeding 3000 AADT, different parking lane widths, parking turnovers and one road was multilane. Annerley Road was the most constrained road environment with 2.5m parking lane and 3.3m traffic lane and had the highest AADT. In peak and non-peak conditions, the majority of motorists chose to follow rather than pass a cyclist. Queuing was observed in peak conditions. The impact of the queuing on the traffic stream was minimal due to the interspersion of BAZ and bicycle lanes. The TRUM recommends that BAZ could be used on roads with a minimum 2.5m parking lane and a minimum 3m travel lane. Based on the recorded operating envelope, motorists would need to cross the dividing line to pass a cyclist in this arrangement if there was active on-street parking. Designers should be mindful of the parameters affecting the safety of BAZ in narrow conditions including parking demand, traffic volume and passing opportunities. At Riding Road with a 3m parking lane and 3.3m traffic lane, the majority of motorists passed rather than followed the cyclist. The operating envelopes for Annerley Road and Riding Road were relatively consistent at approximately 1.55m for peak conditions and 1.85m for nonpeak conditions. Queuing was minimal even in peak hour. On Kedron Brook Road, where the parking lane was 3.5m and the traffic lane was 3.3m, all motorists had sufficient width to pass a cyclist. Consistently on all three roads, cyclists rode on about a 3m offset to the kerb line when passing a parked car, and there could be a case to argue that the centre of the BAZ symbol would be better placed 3m off the kerb on roads with high parking demand, rather than on the parking edge line. In these instances where the symbol could be up to 0.5m from the edge line, a different bike symbol has been developed that could be used to give a better indication of the travel path of cyclists. Another advantage of the 3 m offset would be that the symbol would not be obscured by a parked car which is a problem with narrow parking lanes. When a cyclist was not passing a parked car, they were observed to ride to the left of the edge line. Thus, for roads with low parking demand and narrow 2.5m parking lanes, the symbol is probably best left placed centrally on the edge line to deter cyclists from unnecessarily riding in the traffic lane. Nicole Eileen Smart 53

70 Another option that could be considered for the marking of the BAZ symbol could be the implementation of wider symbols. If the centre of the symbol is supposed to be the desired travel path for cyclists (given the width of the symbol is only 1.1m wide and the minimum envelope is 1m), then the standardized 1530mm symbol gives some latitude in the cyclists alignment. The symbol would also be more visible. Operating envelope data was not obtained for roads without parking lanes, but using the offset to the parking lane as the expected offset to a kerb line, the placement of the BAZ symbol for roads without parking seems to be appropriate. 7.3 BAZ Questionnaire Survey Separate questions were developed for motorists and members of bicycle user groups. The response from the motorists clearly showed a lack of understanding of the purpose of BAZ and less than a quarter were aware that BAZ indicated the position cyclists are expected to ride. Slightly more than half of motorists recognized the value of BAZ to warn of the presence of cyclists. Almost all respondents recognized that the white symbol designated a bicycle lane. The lack of understanding of BAZ linemarking, particularly by motorists, is a concern and warrants an education campaign. The responses from cyclists indicated a greater understanding of bicycle facilities. As would be expected, cyclists commented that they feel safer riding in a bicycle lane than a BAZ. However, BAZ were still considered to be safe by cyclists. A comparison was undertaken of responses given to a similar questionnaire by Connell Wagner Awareness has greatly increased and understanding has improved since the Connell Wagner study. 7.4 Crash Data Analysis An analysis was undertaken of 1100 Brisbane cyclist crashes for the period 2004 to 2009 to determine if BAZ facilities are any more dangerous than dedicated bicycle lanes. The majority of crashes occurred on roads without bicycle facilities and as such, only 100 crashes were analysed. Crash rate calculations, based on crashes per km per year of each type of facility, suggest that cycling in a BAZ is no more hazardous than cycling in a designated 54 An Evaluation of Bicycle Awareness Zones

71 bicycle lane. With only a small dataset, it is difficult to draw any succinct conclusions but it would appear that mid-block crashes are more likely to occur on BAZ due to greater probability of dooring. Five dooring incidents were recorded and these all occurred where the combined width of the parking lane and adjoining traffic lane was 5.9m or less. Of the 12 mid-block crashes which were associated with a BAZ, it is possible that eight of these crashes could have been avoided if there had been sufficient width to provide a dedicated bicycle lane. The majority of BAZ (70% of total) and bicycle lane (89% of total) crashes occurred at intersections which are known to be regions of high potential conflict. Aside from the greater risk of a dooring incident (which was only minimal over the five years), the available data suggests that BAZ don t have any obvious crash history that would lead to a conclusion that BAZ are unsafe to use. 7.5 Summary Video analysis demonstrated that BAZ can be effective on roads with up to AADT, including multilane roads, with minimal impact to the traffic flow. In constrained width conditions, the impact on the traffic stream can be minimized by interspersing BAZ with bicycle lanes; changes to the BAZ symbol, position and/ or size, could be justified based on calculated operating envelopes and offsets from parked cars; Questionnaire survey showed motorists had a lack of understanding of the purpose of the BAZ symbols. Cyclists had a greater understanding. Both motorists and cyclists groups agreed that BAZ makes a safer road environment for cyclists than a road without any bicycle symbols at all; based on the outcomes of the crash analysis, there is no evidence to suggest BAZ are statistically more dangerous than a bicycle lane; bicycle lanes should always be used in preference to a BAZ where the carriageway width permits but BAZ have proven to be a useful and effective link in local government cycle networks. Nicole Eileen Smart 55

72 8 RECOMMENDATIONS The outcomes of the project have helped to define a number of recommendations which will lead to an increase in utilization of BAZ and improvements to the safety and effectiveness of BAZ. The recommendations are as follows: more research to confirm the findings of this project which indicated that BAZ can be effective on roads with up to AADT, including multilane roads, subject to consideration of available lane widths, cyclists operating envelopes, passing opportunities, and parking demand with minimal impact to the traffic flow. ensure passing opportunities are provided on BAZ routes with heavy traffic volumes to minimize the potential for queuing; adopt the standard 1530mm BAZ symbol to better reflect the operating envelopes; consider the use of tracking lines with the larger BAZ symbol; relocate the BAZ symbol 3m from the kerb. For roads with parking lanes less than 3m and where the travel lane is 3.3m or less, the symbol should only be offset off the edge line where parking demand is high; reduce the spacing between BAZ symbols and modify the design of the symbol; delineate parking bays on roads with a wide parking lane (3m or wider); install double continuity lines through intersections with BAZ to increase passing opportunities; the TRUM should include the cyclist envelope data for the various road width conditions to assist with the design of cycle facilities; educate motorists and cyclists about BAZ including the purposes of the various bicycle facilities; how motorists should drive in the presence of a cyclist in the various facilities; inform motorists of the reasons cyclists may need to cycle further in the traffic lane; and also advise cyclists of the ideal position they should be riding in a BAZ; local governments should document the date and locations of each BAZ and bicycle lane installed; regular counts of the AADT and cyclist numbers along cycle routes. 56 An Evaluation of Bicycle Awareness Zones

73 9 REFERENCES Alta Planning and Design 2004, San Francisco s Shared Lane Pavement Markings: Improving Bicycle Safety, San Francisco. Amy Gillett Foundation 2009, A Metre Matters, viewed 5/6/2011, < Austroads 1999, Guide to Traffic Engineering Practice, Part 14: Bicycles, AP-11.14/99, Austroads, Sydney. Austroads 2011, Cycling Aspects of Austroads Guides, AP-G88/11, Austroads, Sydney. Bhat, C, Eluru, N, Hallett, I, Hlavacek, I, Karl, A, Machemehl, R, Sener, I, Torrance, K The Effects of On-Street Parking on Cyclist Route Choice and the Operational Behaviour of Cyclists and Motorists, Centre for Transportation Research at The University of Texas at Austin, USA. Connell Wagner 2001, Bicycle Brisbane Plan Review: Review of Bicycle- Friendly Zones, CT/ Revision 1, Brisbane. Cycling England 2008, Design Portfolio: Cycle Lanes, viewed 22/3/2011, < Federal Highway Administration 2010, Evaluation of Shared Lane Markings, Publication Number FHWA- HRT , U.S. Department of Transportation. Forester, J 1997, Bicycle Transportation: A Handbook for Cycling Transportation Engineers, The MIT Press, Massachusetts. Houten, R & Seiderman, C 2005, How Pavement Markings Influence Bicycle and Motor Vehicle Positioning Case Study in Cambridge, Massachusetts, Journal of the Transportation Research Board, no.1939, p. 3-14, viewed 4 March 2011, via TMR Library Services Catalogue. Nicole Eileen Smart 57

74 National Committee on Uniform Traffic Control Devices 2005, Minutes from 7/1/2005, Arizona. Queensland Government 2010, Data Analysis Road Crash Glossary, Brisbane. Rakotonirainy, A & Schramm, A 2010, Lane Width: The effect on the safety of cyclists in urban areas, Journal of the Australasian College of Road Safety, vol. 21 no. 2, pp RTA 2005, NSW Bicycle Guidelines, , RTA, NSW. Traffic Management Practices Committee 2011, Minutes from 4/2/2011, Transport and Main Roads, Brisbane. Transport and Main Roads 2006, Road Planning and Design Manual: Chapter 14 Roundabouts, Transport and Main Roads, Brisbane. Transport and Main Roads 2009, Traffic and Road Use Management Manual: Bicycle Awareness Zones, Transport and Main Roads, Brisbane. Yeates, M 2000, Making space for cyclists by sharing the road Brisbane City Council s Bicycle Friendly Zone, paper presented to the National Road Safety Conference, Brisbane, 26 November An Evaluation of Bicycle Awareness Zones

75 APPENDIX A: VIDEO ANALYSIS CODING Appendix A- Video Analysis Coding 59

76 APPENDIX B: SUGGESTED SYMBOL DESIGNS B MM X 1800MM SYMBOL, 3M OFFSET FROM KERB, WITH YELLOW TRACKING LINES 60 An Evaluation of Bicycle Awareness Zones

77 B MM X 1800MM SYMBOL, 3M OFFSET FROM KERB, WITH ARROWS AS TRACKING LINES Appendix B- Suggested Symbol Designs 61

78 B MM X 2500MM SYMBOL, 3M OFFSET FROM KERB, WITH YELLOW TRACKING LINES 62 An Evaluation of Bicycle Awareness Zones

79 B MM X 2500MM SYMBOL, 3M OFFSET FROM KERB, WITH ARROWS AS TRACKING LINES Appendix B- Suggested Symbol Designs 63

80 B MM X 2500MM SYMBOL, 3M OFFSET FROM KERB, NO TRACKING LINES 64 An Evaluation of Bicycle Awareness Zones

81 APPENDIX C: TRUM GUIDELINES C.1 LANE CONFIGURATIONS WITHOUT PARKING (Transport and Main Roads 2009) Appendix C- TRUM Guidelines 65

82 C.2 ROAD CONFIGURATIONS WITH PARALLEL PARKING (Transport and Main Roads 2009) 66 An Evaluation of Bicycle Awareness Zones

83 APPENDIX D: D.1 MOTORISTS SURVEY BAZ QUESTIONNAIRE SURVEYS Appendix D- BAZ Questionnaire Surveys 67

84 68 An Evaluation of Bicycle Awareness Zones

85 Appendix D- BAZ Questionnaire Surveys 69

86 70 An Evaluation of Bicycle Awareness Zones

87 Appendix D- BAZ Questionnaire Surveys 71

88 72 An Evaluation of Bicycle Awareness Zones

89 Appendix D- BAZ Questionnaire Surveys 73

90 74 An Evaluation of Bicycle Awareness Zones

91 Appendix D- BAZ Questionnaire Surveys 75

92 76 An Evaluation of Bicycle Awareness Zones

93 Appendix D- BAZ Questionnaire Surveys 77

94 78 An Evaluation of Bicycle Awareness Zones

95 D.2 CYCLISTS SURVEY Appendix D- BAZ Questionnaire Surveys 79

96 80 An Evaluation of Bicycle Awareness Zones

97 Appendix D- BAZ Questionnaire Surveys 81

98 82 An Evaluation of Bicycle Awareness Zones

99 Appendix D- BAZ Questionnaire Surveys 83

100 84 An Evaluation of Bicycle Awareness Zones

101 Appendix D- BAZ Questionnaire Surveys 85

102 86 An Evaluation of Bicycle Awareness Zones

103 Appendix D- BAZ Questionnaire Surveys 87